output
stringlengths 7
3.46k
| input
stringclasses 1
value | instruction
stringlengths 129
114k
|
|---|---|---|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: One or more techniques and/or systems are provided for training and/or utilizing a traffic obstruction identification model for identifying traffic obstructions based upon vehicle location point data. For example, a training dataset, comprising sample vehicle location points (e.g., global positioning system location points of vehicles) and traffic obstruction identification labels (e.g., locations of known traffic obstructions such as stop signs, crosswalks, stop lights, etc.), may be evaluated to extract a set of training features indicative of traffic flow patterns. The set of training features and the traffic obstruction identification labels may be used to train a traffic obstruction identification model to create a trained traffic obstruction identification model. The trained traffic obstruction identification model may be used to determine whether a road segment has a traffic obstruction or not. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: One or more techniques and/or systems are provided for training and/or utilizing a traffic obstruction identification model for identifying traffic obstructions based upon vehicle location point data. For example, a training dataset, comprising sample vehicle location points (e.g., global positioning system location points of vehicles) and traffic obstruction identification labels (e.g., locations of known traffic obstructions such as stop signs, crosswalks, stop lights, etc.), may be evaluated to extract a set of training features indicative of traffic flow patterns. The set of training features and the traffic obstruction identification labels may be used to train a traffic obstruction identification model to create a trained traffic obstruction identification model. The trained traffic obstruction identification model may be used to determine whether a road segment has a traffic obstruction or not. |
|
Certain ensembles of metapletic anyons allow for topologically protected encoding and processing of quantum information. Such processing is done by sequences of gates (circuits) drawn from a certain basis of unitary metaplectic gates. A subject unitary operator required for the desired processing can be approximated to any desired precision by a circuit that has to be effectively and efficiently synthesized on a classical computer. Synthesis methods use unitary reflection operators that can be represented either exactly or by ancilla-assisted approximation over the basis of metaplectic gates based on cost-optimizing determinations made by the synthesis algorithm. | Please help me write a proper abstract based on the patent claims. CLAIM: 1-15. (canceled) 16. A computer-implemented synthesis method for quantum computing, comprising: with a processor: selecting a plurality of weakly-integral anyons; and defining an asymptotically optimal quantum circuit approximating a selected unitary operator to a desired precision with a series of gates to be applied to the selected plurality of weakly-integral anyons. 17. The method of claim 16, wherein the weakly-integral anyons are metaplectic anyons. 18. The method of claim 17, wherein gates of the series of gates in the quantum circuit are selected from an extended metaplectic basis. 19. The method of claim 17, wherein gates of the series of gates in the quantum circuit are selected from an augmented metaplectic basis 20. The method of claim 19, wherein the series of gates includes a set of gates corresponding to an axial reflection operator. 21. The method of claim 20, wherein the axial reflection operator is an n-qutrit axial reflection operator, with a recursively-built exact representation in terms of single-qutrit INC and Flip gates and two-qutrit SWAP and SUM gates. 22. The method of claim 20, wherein if the axial reflection operator is an n-qutrit axial reflection operator for n>5, then n-qutrit axial reflection operator is synthesized as an approximate n-qutrit reflection operator using (n−1) ancillary qutrits. 23. The method of claim 20 wherein the selected unitary is a two-level multi-qutrit unitary, and further comprising generating an intermediate representation as a product of two-level reflections and, based on the intermediate representation, defining the quantum circuit as a series of metaplectic gates and axial reflections. 24. The method of claim 20, further comprising representing the selected unitary as a product of two-level unitaries, and defining the asymptotically optimal quantum circuit based on the representation as a product of two-level unitaries. 25. The method of claim 20, wherein the methods of claims 8 and 9 are applied to a known exact decomposition of the selected multi-qutrit unitary into a product of several two-level unitaries and at most one diagonal unitary. 26. The method of claim 20, further comprising determining whether to define the asymptotically optimal quantum circuit based on exact axial reflections or as an ancilla-assisted circuit based on approximated reflections. 27. A circuit synthesis tool, comprising: at least one processor; and a memory storing processor-executable instructions for synthesizing a unitary over a metaplectic basis. 28. The circuit synthesis tool of claim 27, wherein the memory stores processor-executable instructions for representing a qutrit with coefficients that are Eisenstein rationals. 29. The circuit synthesis tool of claim 27, further comprising determining whether to define the asymptotically optimal quantum circuit based on exact axial reflections or as an ancilla-assisted circuit based on approximated reflections. 30. The circuit synthesis tool of claim 29, wherein the memory stores processor-executable instructions for synthesizing an n-qutrit circuit using n-ancillas. 31. The circuit synthesis tool of claim 30, wherein the memory stores processor-executable instructions for synthesizing the unitary using at least one two-qutrit entangler. 32. The circuit synthesis tool of claim 27, wherein the memory stores processor-executable instructions for determining if the unitary is to be synthesized without ancillas, and synthesizing the circuit without ancillas based on the determination. 33. The circuit synthesis tool of claim 32, wherein the memory stores processor-executable instructions for synthesizing the circuit so as to include one or more SUM, SWAP, and axial reflection gates. 34. The circuit synthesis tool of claim 27, wherein the circuit is synthesized with respect to at least one qudit. 35. A computer implemented quantum circuit synthesis method, comprising: defining a universal metaplectic basis; based on a precision and a number of qutrits, determining whether a selected unitary is to be implemented with ancillas or without ancillas; if synthesis without ancillas is selected, synthesizing the unitary based on a series of SUM, SWAP, and axial reflection gates, wherein the axial reflection gates are represented exactly based on a single-qutrit axial reflection gates; and if synthesis with ancillas is selected and the unitary is an n-qutrit unitary, synthesizing the n-qutrit unitary approximately using n ancillas. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Certain ensembles of metapletic anyons allow for topologically protected encoding and processing of quantum information. Such processing is done by sequences of gates (circuits) drawn from a certain basis of unitary metaplectic gates. A subject unitary operator required for the desired processing can be approximated to any desired precision by a circuit that has to be effectively and efficiently synthesized on a classical computer. Synthesis methods use unitary reflection operators that can be represented either exactly or by ancilla-assisted approximation over the basis of metaplectic gates based on cost-optimizing determinations made by the synthesis algorithm. |
|
G06N99002 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Certain ensembles of metapletic anyons allow for topologically protected encoding and processing of quantum information. Such processing is done by sequences of gates (circuits) drawn from a certain basis of unitary metaplectic gates. A subject unitary operator required for the desired processing can be approximated to any desired precision by a circuit that has to be effectively and efficiently synthesized on a classical computer. Synthesis methods use unitary reflection operators that can be represented either exactly or by ancilla-assisted approximation over the basis of metaplectic gates based on cost-optimizing determinations made by the synthesis algorithm. |
|
Method and Apparatus for rapid scalable unified infrastructure system management platform are disclosed by discovery of compute nodes, network components across data centers, both public and private for a user; assessment of type, capability, VLAN, security, virtualization configuration of the discovered unified infrastructure nodes and components; configuration of nodes and components covering add, delete, modify, scale; and rapid roll out of nodes and components across data centers both public and private. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method, comprising: evaluations of resource pool usage patterns and burn rates; and intelligent forecasting of resource pool exhaustion 2. An apparatus, comprising: a software based forecast of cloud resource pool exhaustion; and a software platform that evaluates resource usage patterns and burn rates |
|
REJECTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Method and Apparatus for rapid scalable unified infrastructure system management platform are disclosed by discovery of compute nodes, network components across data centers, both public and private for a user; assessment of type, capability, VLAN, security, virtualization configuration of the discovered unified infrastructure nodes and components; configuration of nodes and components covering add, delete, modify, scale; and rapid roll out of nodes and components across data centers both public and private. |
|
G06N502 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Method and Apparatus for rapid scalable unified infrastructure system management platform are disclosed by discovery of compute nodes, network components across data centers, both public and private for a user; assessment of type, capability, VLAN, security, virtualization configuration of the discovered unified infrastructure nodes and components; configuration of nodes and components covering add, delete, modify, scale; and rapid roll out of nodes and components across data centers both public and private. |
|
A computer system access in a database a first linear sequence table including a plurality of entries. A respective entry of the plurality of entries includes sequential state information for a respective user. The sequential state information for the respective entry identifies a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time. The computer system initiates aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for processing big data, comprising: at a computer system with one or more processors and memory: accessing in a database a first linear sequence table including a plurality of entries, wherein a respective entry of the plurality of entries includes sequential state information for a respective user, the sequential state information for the respective entry identifying a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time; and initiating aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. 2. The method of claim 1, wherein: aggregation of data in the first linear sequence table includes grouping and/or counting entries that are associated with the particular preceding event and the particular subsequent event. 3. The method of claim 1, comprising: accessing in a database a first table including a plurality of entries, wherein: a respective entry of the plurality of entries includes state information and sequence information for a respective user, the state information for the respective entry identifying a respective event associated with the respective user and the sequence information for the respective entry identifying a sequence of the respective event within a plurality of events associated with the respective user; and the plurality of entries includes multiple entries for the respective user; accessing in the database a second table that corresponds to the first table; and filling the first linear sequence table based on entries in the first table and the second table. 4. The method of claim 3, wherein the first linear sequence table is formed in response to a single instruction. 5. The method of claim 3, wherein the second table is identical to the first table or the second table is a mirror image of the first table. 6. The method of claim 3, wherein the first table includes information identifying respective users; the second table includes information identifying the respective users; and the first linear sequence table does not include information identifying the respective users. 7. The method of claim 3, wherein the first linear sequence table does not include the sequence information. 8. The method of claim 3, wherein the first table includes a first number of entries for the respective user and the first linear sequence table includes a second number of entries for the respective user that is distinct from the first number. 9. The method of claim 3, further comprising forming the first linear sequence table. 10. The method of claim 1, further comprising: obtaining respective quantities corresponding to respective numbers of entries that are associated with respective subsequent events and one or more preceding events; and selecting, for the one or more preceding events, a subsequent event based on a quantity that corresponds to a number of entries that are associated with the one or more preceding events and the selected subsequent event. 11. The method of claim 1, further comprising: obtaining respective quantities corresponding to respective numbers of entries that are associated with respective preceding events and one or more subsequent events; and selecting, for the one or more subsequent events, a preceding event based on a quantity that corresponds to a number of entries that are associated with the one or more subsequent events and the selected preceding event. 12. The method of claim 1, further comprising: obtaining respective quantities corresponding to respective numbers of entries that are associated with respective subsequent events and a first preceding event; selecting, for the first preceding event, a first event based on a quantity that corresponds to a number of entries that are associated with the first preceding event and the first event as a subsequent event; obtaining respective quantities corresponding to respective numbers of entries that are associated with respective subsequent events and a set of the first preceding event and the first event as preceding events; and selecting, for the set of the first preceding event and the first event, a second event based on a quantity that corresponds to a number of entries that are associated with the set of the first preceding event and the first event as preceding events and the second event as a subsequent event. 13. The method of claim 12, further comprising: obtaining respective quantities corresponding to respective numbers of entries that are associated with respective subsequent events and a set of the first preceding event, the first event, and the second event as preceding events; and selecting, for the set of the first preceding event, the first event, and the second event, a third event based on a quantity that corresponds to a number of entries that are associated with the set of the first preceding event, the first event, and the second event as preceding events, and the third event as a subsequent event. 14. The method of claim 1, further comprising: filling a first multi-dimensional sequence table, wherein: one of a column and a row of the first multi-dimensional sequence table corresponds to the preceding events; the other one of the column and the row of the first multi-dimensional sequence table corresponds to the subsequent events; and an entry in the first multi-dimensional sequence table includes a quantity that corresponds to a number of entries that correspond to a respective preceding event and a respective subsequent event of the first linear sequence table. 15. The method of claim 14, further comprising: accessing a second multi-dimensional sequence table, wherein: a column of the second multi-dimensional sequence table corresponds to the column of the first multi-dimensional sequence table; a row of the second multi-dimensional sequence table corresponds to the row of the first multi-dimensional sequence table; an entry in the second multi-dimensional sequence table includes a quantity that corresponds to a number of entries that correspond to a respective preceding event and a respective subsequent event; and the second multi-dimensional sequence table is distinct from the first multi-dimensional sequence table; and obtaining respective quantities corresponding to respective numbers of entries, in the first multi-dimensional sequence table, that are associated with a first set of one or more preceding events; obtaining respective quantities corresponding to respective numbers of entries, in the second multi-dimensional sequence table, that are associated with a second set of one or more preceding events; and selecting, collectively for the first set of one or more preceding events for the first multi-dimensional sequence table and for the second set of one or more preceding events for the second multi-dimensional sequence table, a particular subsequence event based on the respective quantities corresponding to the respective numbers of entries, in the first multi-dimensional sequence table, that are associated with the first set of one or more preceding events and the respective quantities corresponding to the respective numbers of entries, in the second multi-dimensional sequence table, that are associated with the second set of one or more preceding events. 16. The method of claim 1, further comprising: accessing in the database a second linear sequence table including a plurality of entries, wherein a respective entry of the plurality of entries includes sequential state information for a respective user, the sequential state information for the respective entry identifying a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time; initiating aggregation of data in the second linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries; obtaining respective quantities corresponding to respective numbers of entries, in the first linear sequence table, that are associated with a first set of one or more preceding events; obtaining respective quantities corresponding to respective numbers of entries, in the second linear sequence table, that are associated with a second set of one or more preceding events; and selecting, collectively for the first set of one or more preceding events for the first linear sequence table and for the second set of one or more preceding events for the second linear sequence table, a particular subsequent event based on the respective quantities corresponding to the respective numbers of entries, in the first linear sequence table, that are associated with the first set of one or more preceding events and the respective quantities corresponding to the respective numbers of entries, in the second linear sequence table, that are associated with the second set of one or more preceding events. 17. A computer system, comprising: one or more processors; and memory storing one or more programs, which, when executed by the one or more processors, cause the computer system to: access in a database a first linear sequence table including a plurality of entries, wherein a respective entry of the plurality of entries includes sequential state information for a respective user, the sequential state information for the respective entry identifying a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time; and initiate aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. 18. A computer readable storage medium, storing one or more programs for execution by one or more processors of a computer system, the one or more programs including instructions for: accessing in a database a first linear sequence table including a plurality of entries, wherein a respective entry of the plurality of entries includes sequential state information for a respective user, the sequential state information for the respective entry identifying a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time; and initiating aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A computer system access in a database a first linear sequence table including a plurality of entries. A respective entry of the plurality of entries includes sequential state information for a respective user. The sequential state information for the respective entry identifies a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time. The computer system initiates aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A computer system access in a database a first linear sequence table including a plurality of entries. A respective entry of the plurality of entries includes sequential state information for a respective user. The sequential state information for the respective entry identifies a respective preceding event associated with a respective preceding time and a respective subsequent event associated with a respective subsequent time that is subsequent to the respective preceding time. The computer system initiates aggregation of data in the first linear sequence table to obtain a quantity that corresponds to a number of entries that are associated with a particular preceding event and a particular subsequent event of preceding events and subsequent events of the plurality of entries. |
|
Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using recurrent neural networks to analyze health events. One of the methods includes: processing each of a plurality of initial temporal sequences of health events to generate, for each of the initial temporal sequences, a respective network internal state of a recurrent neural network for each time step in the initial temporal sequence; storing, for each of the initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in a repository; obtaining a first temporal sequence; processing the first temporal sequence using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and selecting one or more initial temporal sequences that are likely to include health events that are predictive of future health events in the first temporal sequence. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: obtaining a plurality of initial temporal sequences of health events, wherein each of the plurality of initial temporal sequences of health events is associated with a different patient, wherein each of the initial temporal sequences comprises respective health-related data associated with the respective patient that is associated with the initial temporal sequence at each of a plurality of time steps, and wherein, for one or more of the time steps in each of the initial temporal sequences, the health-related data at the time step is a respective token from a predetermined vocabulary of tokens, each token in the vocabulary representing a different health event; processing each of the plurality of initial temporal sequences of health events using a recurrent neural network to generate, for each of the initial temporal sequences, a respective network internal state of the recurrent neural network for each time step in the initial temporal sequence, wherein the recurrent neural network has been trained to receive input temporal sequences and, for each time step in each input temporal sequence, generate a network internal state for the time step and predict future events occurring after the health event identified at the time step from the network internal state for the time step; storing, for each of the plurality of initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in an internal state repository; obtaining a first temporal sequence of health events that is associated with a current patient; processing the first temporal sequence of health events using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and comparing the sequence internal state for the first temporal sequence to the network internal states for the initial temporal sequences that are stored in the internal state repository to determine a plurality of network internal states from the network internal states for the initial temporal sequences that are similar to the sequence internal state; selecting, as temporal sequences that are likely to include health events that are predictive of future health events that may become associated with the current patient, the initial temporal sequences corresponding to the similar network internal states. 2. (canceled) 3. The method of claim 1, wherein determining network internal states in the internal state repository that are similar to the sequence internal state comprises: computing, for each of the network internal states in the internal state repository, a respective similarity measure between the network internal state and the sequence internal state; and determining the similar network internal states from the similarity measures. 4. The method of claim 1, further comprising: associating each network internal state in the internal state repository with a respective time step and a respective initial temporal sequence for which the network internal state was generated. 5. The method of claim 4, further comprising: providing, for presentation to a user, data identifying, for each of the selected initial temporal sequences, the health events in the selected initial temporal sequences that are at time steps subsequent to the time step for which the corresponding network internal state was generated. 6. The method of claim 4, further comprising: computing, from the health events in the selected initial temporal sequences that are at time steps subsequent to the time step for which the corresponding network internal state was generated, a statistic for a particular health event identifying a frequency of occurrence of the particular health event; and providing the computed statistic for presentation to a user. 7. The method of claim 1, wherein the recurrent neural network is trained to generate, for each of the plurality of time steps in each input training sequence, a respective score for each of a plurality of possible health events from the network internal state for the time step, wherein the respective score for each of the possible health events represents a likelihood that the possible health event is a health event at a time step subsequent to the time step in the input training sequence. 8. The method of claim 1, wherein processing the first temporal sequence of health events using the recurrent neural network to generate a sequence internal state for the first temporal sequence comprises: for each time step in the first temporal sequence, processing the data identifying the health event for the time step using the recurrent neural network to generate a network internal state for the time step; and selecting the network internal state for a last time step in the first temporal sequence as the sequence internal state for the first temporal sequence. 9. (canceled) 10. (canceled) 11. The method of claim 1, wherein, for one or more of the time steps in each of the initial temporal sequences, the health-related data at the time step is other health-related data classified as impacting the health of the respective patient. 12. A system comprising one or more computers and one or more storage devices storing instructions that when executed by the one or more computers cause the one or more computers to perform operations comprising: obtaining a plurality of initial temporal sequences of health events, wherein each of the plurality of initial temporal sequences of health events is associated with a different patient, wherein each of the initial temporal sequences comprises respective health-related data associated with the respective patient that is associated with the initial temporal sequence at each of a plurality of time steps, and wherein, for one or more of the time steps in each of the initial temporal sequences, the health-related data at the time step is a respective token from a predetermined vocabulary of tokens, each token in the vocabulary representing a different health event; processing each of the plurality of initial temporal sequences of health events using a recurrent neural network to generate, for each of the initial temporal sequences, a respective network internal state of the recurrent neural network for each time step in the initial temporal sequence, wherein the recurrent neural network has been trained to receive input temporal sequences and, for each time step in each input temporal sequence, generate a network internal state for the time step and predict future events occurring after the health event identified at the time step from the network internal state for the time step; storing, for each of the plurality of initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in an internal state repository; obtaining a first temporal sequence of health events that is associated with a current patient; processing the first temporal sequence of health events using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and comparing the sequence internal state for the first temporal sequence to the network internal states for the initial temporal sequences that are stored in the internal state repository to determine a plurality of network internal states from the network internal states for the initial temporal sequences that are similar to the sequence internal state; selecting, as temporal sequences that are likely to include health events that are predictive of future health events that may become associated with the current patient, the initial temporal sequences corresponding to the similar network internal states. 13. (canceled) 14. The system claim 12, wherein determining network internal states in the internal state repository that are similar to the sequence internal state comprises: computing, for each of the network internal states in the internal state repository, a respective similarity measure between the network internal state and the sequence internal state; and determining the similar network internal states from the similarity measures. 15. The system of claim 12, the operations further comprising: associating each network internal state in the internal state repository with a respective time step and a respective initial temporal sequence for which the network internal state was generated. 16. The system of claim 15, the operations further comprising: providing, for presentation to a user, data identifying, for each of the selected initial temporal sequences, the health events in the selected initial temporal sequences that are at time steps subsequent to the time step for which the corresponding network internal state was generated. 17. The system of claim 15, the operations further comprising: computing, from the health events in the selected initial temporal sequences that are at time steps subsequent to the time step for which the corresponding network internal state was generated, a statistic for a particular health event identifying a frequency of occurrence of the particular health event; and providing the computed statistic for presentation to a user. 18. The system of claim 12, wherein the recurrent neural network is trained to generate, for each of the plurality of time steps in each input training sequence, a respective score for each of a plurality of possible health events from the network internal state for the time step, wherein the respective score for each of the possible health events represents a likelihood that the possible health event is a health event at a time step subsequent to the time step in the input training sequence. 19. The system of claim 12, wherein processing the first temporal sequence of health events using the recurrent neural network to generate a sequence internal state for the first temporal sequence comprises: for each time step in the first temporal sequence, processing the data identifying the health event for the time step using the recurrent neural network to generate a network internal state for the time step; and selecting the network internal state for a last time step in the first temporal sequence as the sequence internal state for the first temporal sequence. 20. A computer program product encoded on one or more non-transitory computer readable media, the computer program product comprising instructions that when executed by one or more computers cause the one or more computers to perform operations comprising: obtaining a plurality of initial temporal sequences of health events, wherein each of the plurality of initial temporal sequences of health events is associated with a different patient, wherein each of the initial temporal sequences comprises respective health-related data associated with the respective patient that is associated with the initial temporal sequence at each of a plurality of time steps, and wherein, for one or more of the time steps in each of the initial temporal sequences, the health-related data at the time step is a respective token from a predetermined vocabulary of tokens, each token in the vocabulary representing a different health event; processing each of the plurality of initial temporal sequences of health events using a recurrent neural network to generate, for each of the initial temporal sequences, a respective network internal state of the recurrent neural network for each time step in the initial temporal sequence, wherein the recurrent neural network has been trained to receive input temporal sequences and, for each time step in each input temporal sequence, generate a network internal state for the time step and predict future events occurring after the health event identified at the time step from the network internal state for the time step; storing, for each of the plurality of initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in an internal state repository; obtaining a first temporal sequence of health events that is associated with a current patient; processing the first temporal sequence of health events using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and comparing the sequence internal state for the first temporal sequence to the network internal states for the initial temporal sequences that are stored in the internal state repository from the network internal states for the initial temporal sequences to determine a plurality of network internal states that are similar to the sequence internal state; selecting, as temporal sequences that are likely to include health events that are predictive of future health events that may become associated with the current patient, the initial temporal sequences corresponding to the similar network internal states. 21. The system of claim 12, wherein, for one or more of the time steps in each of the initial temporal sequences, the health-related data at the time step is other health-related data classified as impacting the health of the respective patient. 22. The computer program product of claim 20, wherein selecting one or more initial temporal sequences comprises: determining network internal states in the internal state repository that are similar to the sequence internal state; and selecting the initial temporal sequences for which the similar network internal states were generated as the initial temporal sequences from the plurality of initial temporal sequences that are likely to include health events that are predictive of future health events in the first temporal sequence. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using recurrent neural networks to analyze health events. One of the methods includes: processing each of a plurality of initial temporal sequences of health events to generate, for each of the initial temporal sequences, a respective network internal state of a recurrent neural network for each time step in the initial temporal sequence; storing, for each of the initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in a repository; obtaining a first temporal sequence; processing the first temporal sequence using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and selecting one or more initial temporal sequences that are likely to include health events that are predictive of future health events in the first temporal sequence. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Methods, systems, and apparatus, including computer programs encoded on computer storage media, for using recurrent neural networks to analyze health events. One of the methods includes: processing each of a plurality of initial temporal sequences of health events to generate, for each of the initial temporal sequences, a respective network internal state of a recurrent neural network for each time step in the initial temporal sequence; storing, for each of the initial temporal sequences, one or more of the network internal states for the time steps in the temporal sequence in a repository; obtaining a first temporal sequence; processing the first temporal sequence using the recurrent neural network to generate a sequence internal state for the first temporal sequence; and selecting one or more initial temporal sequences that are likely to include health events that are predictive of future health events in the first temporal sequence. |
|
The present invention relates to systems and methods for automated troubleshooting. User identification is recorded, and the problem to be solved is determined. A series of decision trees are used to guide the user through troubleshooting. If the problem is resolved at any point, the event is logged as successful. This logging includes a listing of the steps taken by the user. The log may be employed to tune the decision tree for more optimal performance in the future. If no successful solutions are achieved, then the session may be forwarded to a human representative for resolution. Part of this forwarding includes classifying the event by at least one failure code. The failure code reflects the steps taken while guiding the user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for automated troubleshooting comprising: recording user identification; determining a problem to be solved; assessing the user's technical competency level; guiding the user through a decision tree using an automated system with a processor, wherein the decision tree is based upon the user's technical competency level; and recording steps taken, while guiding the user, as an event. 2. The method of claim 1, further comprising resolving the problem and logging the event as successful, wherein the logging includes a listing of the steps taken by the user. 3. The method of claim 2, wherein the listing of the steps taken by the user is used to tune the decision tree. 4. The method of claim 1, further comprising elevating the user to a higher decision tree. 5. The method of claim 4, further comprising exhausting the decision tree and the higher decision tree without resolving the problem. 6. The method of claim 5, further comprising classifying the event by at least one failure code, wherein the failure code reflects the steps taken while guiding the user. 7. The method of claim 6, further comprising elevating the problem to a human representative. 8. The method of claim 7, further comprising providing the event classification to the human representative. 9. The method of claim 8, wherein human representative is selected by language skills, area of expertise and skill level. 10. The method of claim 1, wherein the guiding the user is done in a language chosen by the user. 11. An automated troubleshooting system comprising: an interface configured to receive user identification; a system diagnostic engine configured to determine a problem to be solved; an automated solutions provider, having a processor, configured to assess the user's technical competency level, and guide the user through a decision tree wherein the decision tree is based upon the user's technical competency level; and a database configured to record steps taken, while guiding the user, as an event. 12. The system of claim 11, wherein the automated solutions provider comprising resolves the problem and wherein the database logs the event as successful, wherein the logging includes a listing of the steps taken by the user. 13. The system of claim 12, wherein the listing of the steps taken by the user is used to tune the decision tree. 14. The system of claim 11, wherein the automated solutions provider elevates the user to a higher decision tree. 15. The system of claim 14, wherein the automated solutions provider exhausts the decision tree and the higher decision tree without resolving the problem. 16. The system of claim 15, wherein the automated solutions provider classifies the event by at least one failure code which is stored in the database, wherein the failure code reflects the steps taken while guiding the user. 17. The system of claim 16, further comprising a human interface configured to elevate the problem to a human representative. 18. The system of claim 17, wherein the human interface provides the event classification to the human representative. 19. The system of claim 18, wherein the human interface selects the human representative based upon language skills, area of expertise and skill level. 20. The system of claim 11, wherein the automated solutions provider guides the user in a language chosen by the user. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: The present invention relates to systems and methods for automated troubleshooting. User identification is recorded, and the problem to be solved is determined. A series of decision trees are used to guide the user through troubleshooting. If the problem is resolved at any point, the event is logged as successful. This logging includes a listing of the steps taken by the user. The log may be employed to tune the decision tree for more optimal performance in the future. If no successful solutions are achieved, then the session may be forwarded to a human representative for resolution. Part of this forwarding includes classifying the event by at least one failure code. The failure code reflects the steps taken while guiding the user. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The present invention relates to systems and methods for automated troubleshooting. User identification is recorded, and the problem to be solved is determined. A series of decision trees are used to guide the user through troubleshooting. If the problem is resolved at any point, the event is logged as successful. This logging includes a listing of the steps taken by the user. The log may be employed to tune the decision tree for more optimal performance in the future. If no successful solutions are achieved, then the session may be forwarded to a human representative for resolution. Part of this forwarding includes classifying the event by at least one failure code. The failure code reflects the steps taken while guiding the user. |
|
A method, procedure and algorithm are provided for efficient resource provisioning in Hadoop MapReduce. The crux of the method, procedure and algorithm is not tied to any specific system, and can be applied to many processes and devices. It provides a general approach and techniques based on an algorithm with mathematical formulas to find the Best Trade-off Point on an elbow curve, non-inverted or inverted, of performance vs. resources. It is applicable to any systems relying on a trade-off elbow curve for making good decision. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method, algorithm, and process for calculating an exact optimal number of resources for a job that provides the best trade-off point between performance vs. resources on an elbow curve ƒ(x)=(a/x)+b fitted from sampled executions of the target system, comprising the steps of: (A) providing a completed configuration and fine tuning of the architecture, software and hardware of a production system targeted for calibration; (B) collecting necessary preview job performance data from historical runtime performances or sampled executions on said the same targeted production system, configured exactly as in step A, as reference points for each workload; (C) curve-fitting said preview job performance data to obtain fit parameters a and b in an elbow curve function ƒ(x)=(a/x)+b, where x is the number of resources; (D) inputting said fit parameter a to a Best-Trade-off-Point algorithm to obtain a recommended optimal number of resources for a workload as output, wherein said Best-Trade-off-Point algorithm includes the steps of: computing the number of resources over a range of slopes from the first derivative of ƒ(x)=(a/x)+b and the acceleration over a range of slopes from the second derivative of ƒ(x)=(a/x)+b; applying the Chain rule to search for break points and major plateaus on the graphs of acceleration, slope, and task resources over a range of incremental changes in acceleration per slope increment; and extracting the exact number of resources at the best trade-off point on the elbow curve and outputs it as recommended optimal number of resources for a workload; (E) repeating steps (B)-(D) gather sufficient resource provisioning data points for different workloads to build a database of resource consumption signatures for subsequent job profiling; (F) repeating steps (A)-(E) to recalibrate said database of resource consumption signatures if there are any major changes to step A; and (G) using said database of resource consumption signatures to match dynamically submitted production jobs to their recommended optimal number of resources for efficient resource provisioning. 2. A method, algorithm, and process for calculating an exact optimal number of resources for a job that provides the best trade-off point between performance vs. resources on an inverted elbow curve ƒ(x)=−(a/x)+b fitted from sampled executions of the target system, comprising the steps of: (A) providing a completed configuration and fine tuning of the architecture, software and hardware of a production system targeted for calibration; (B) collecting necessary preview job performance data from historical runtime performances or sampled executions on said the same targeted production system, configured exactly as in step A, as reference points for each workload; (C) curve-fitting said preview job performance data to obtain fit parameters a and b in an inverted elbow curve function ƒ(x)=−(a/x)+b, where x is the number of resources; (D) inputting said fit parameter a to a Best-Trade-off-Point algorithm to obtain a recommended optimal number of resources for a workload as output, wherein said Best-Trade-off-Point algorithm includes the steps of: computing the number of resources over a range of slopes from the first derivative of ƒ(x)=−(a/x)+b and the acceleration over a range of slopes from the second derivative of ƒ(x)=−(a/x)+b; applying the Chain rule to search for break points and major plateaus on the graphs of acceleration, slope, and task resources over a range of incremental changes in acceleration per slope increment; and extracting the exact number of resources at the best trade-off point on the inverted elbow curve and outputs it as recommended optimal number of resources for a workload; (E) repeating steps (B)-(D) gather sufficient resource provisioning data points for different workloads to build a database of resource consumption signatures for subsequent job profiling; (F) repeating steps (A)-(E) to recalibrate said database of resource consumption signatures if there are any major changes to step A; and (G) using said database of resource consumption signatures to match dynamically submitted production jobs to their recommended optimal number of resources for efficient resource provisioning. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method, procedure and algorithm are provided for efficient resource provisioning in Hadoop MapReduce. The crux of the method, procedure and algorithm is not tied to any specific system, and can be applied to many processes and devices. It provides a general approach and techniques based on an algorithm with mathematical formulas to find the Best Trade-off Point on an elbow curve, non-inverted or inverted, of performance vs. resources. It is applicable to any systems relying on a trade-off elbow curve for making good decision. |
|
G06N5046 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method, procedure and algorithm are provided for efficient resource provisioning in Hadoop MapReduce. The crux of the method, procedure and algorithm is not tied to any specific system, and can be applied to many processes and devices. It provides a general approach and techniques based on an algorithm with mathematical formulas to find the Best Trade-off Point on an elbow curve, non-inverted or inverted, of performance vs. resources. It is applicable to any systems relying on a trade-off elbow curve for making good decision. |
|
The present disclosure relates to the electronic document review field and, more particularly, to various apparatuses and methods of implementing batch-mode active learning for technology-assisted review (TAR) of documents (e.g., legal documents). | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An apparatus configured to implement batch-mode active learning for technology-assisted review (TAR) of documents, the apparatus comprising: a processor; and, a memory that stores processor-executable instructions, wherein the processor interfaces with the memory to execute the processor-executable instructions, whereby the apparatus is operable to: obtain an unlabeled set of documents D; obtain a batch size k; construct a first batch of k documents D; obtain labels for the first batch of k documents D, wherein the labeled first batch of k documents D are referred to as training data documents; construct a classification model, Mc, using the labeled first batch of k documents D; perform an iteration of active learning using the classification model Mc, wherein the perform operation comprises: (i) select a new batch of unlabeled instances Bc using a current version of the classification model, Mc(x), an unlabeled set of available documents D, and the batch size k, wherein the apparatus is further operable to implement a Diversity Sampler process or a Biased Probabilistic Sampler process to select the new batch of unlabeled instances Bc; (ii) obtain labels for the new batch of unlabeled instances Bc; (iii) add the labeled new batch of instances Bc to a current version of the training data documents referred to as extended training data documents Dc; construct an updated classification model M(x) using the extended training data documents Dc; determine whether a stopping criteria has been met; based on the determination that the stopping criteria has not been met, repeat the perform operation, the third construct operation, and the determine operation; and, based on the determination that the stopping criteria has been met, return the updated classification model M(x). 2. The apparatus of claim 1, wherein the classification model, Mc, comprises one of the following: a support vector machine model, a logistic regression model, a nearest neighbors model, decision forest model, neural network model, Bayesian model, or ensemble model. 3. The apparatus of claim 1, wherein the apparatus is further operable to implement the Diversity Sampler process as follows: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), and a similarity threshold t; sort the unlabeled set of available documents D using a sorting function to obtain sorted indices I of the unlabeled set of available documents D; insert the sorted document having a nearest sorted indice I[1] of the sorted unlabeled set of available documents D into the new batch of unlabeled instances Bc; obtain sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; remove the documents with the obtained sorted indices I from the sorted unlabeled set of available documents D; and repeat the insert operation, the second obtain operation, and the remove operation until insert k documents into the new batch of unlabeled instances Bc. 4. The apparatus of claim 3, wherein the sorting function is one of the following: a distance from a hyperplane based on smallest to greatest when the classification model Mc(x) is a support vector machine, SVM, classification model; a distance from a hyperplane based on greatest to smallest when the classification model Mc(x) is a support vector machine, SVM, classification model; a random sorting function; an entropy function; a least confidence function; or a committee disagreement function. 5. The apparatus of claim 1, wherein the apparatus is further operable to implement the Biased Probabilistic Sampler process as follows: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), a similarity threshold t, and a probability distribution function P(x); construct a probability vector w based on an probability distribution function P(x) for each document from the unlabeled set of available documents D; choose a document I[1] from the unlabeled set of available documents D using the weight w of the probability vector; insert the chosen document I[1] into the new batch of unlabeled instances Bc; obtain sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; remove the documents with the obtained sorted indices I from the unlabeled set of available documents D; re-normalize the weight vector as documents have been removed from the sorted unlabeled set of available documents D; and, repeat the choose operation, the insert operation, the second obtain operation, the remove operation, and the re-normalize operation until select the new batch of unlabeled instances Bc. 6. The apparatus of claim 1, wherein the apparatus is further operable to determine whether the stopping criteria has been met by implementing a Kappa agreement which is a measure of how much the current version of the hyperplane hc(x) and the updated hyperplane h(x) agree on respective prediction of labels of a chosen set of the documents D. 7. The apparatus of claim 1, wherein the apparatus is further operable to determine whether the stopping criteria has been met by implementing an Accumulated History of Scores process which comprises the following operations: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, an accumulation function A(Sc, Sc-1, . . . , S1), and a stopping threshold tstop; construct a score vector Sc using the current classification model Mc(x) and the total set of documents DT; combine a current score vector Sc with previous score vectors (Sc-1, . . . , S1) using the accumulation function A(Sc, Sc-1, . . . , S1) to obtain a stability value (s); compare the stability value s to the stopping threshold tstop to determine whether tstop≦S which indicates that the stopping criteria has been met; and based on the determination that the stopping criteria has not been met, store the current score vector Sc to memory as a previous score vector Sc-1. 8. The apparatus of claim 1, wherein the apparatus is further operable to determine whether the stopping criteria has been met by implementing a Cohen's Kappa process which comprises the following operations: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, and a stopping threshold tstop; construct a score vector Sc using the current classification model Mc(x) and the total set of documents DT; retrieve a previous score vector Sc-i; obtain a set of documents D+c from the documents DT that have a positive score in Sc; obtain a set of documents D−c from the documents DT that have a negative score in Sc obtain a set of documents D+(c-1) from the documents DT that have a positive score in Sc-1; obtain a set of documents D−(c-1) from the documents DT that have a negative score in Sc-1; obtain a set of documents in common, D+, between D+c and D+(c-1); obtain a set of documents in common, D−, between D−c and D−(c-1); obtain a probability, P+, of a document having positive score in both score vectors Sc and Sc-1 by counting a number of documents, N+, in D+ divided by the total number of documents, N, in D; obtain a probability, P−, of a document having negative score in both score vectors Sc and Sc-1, by counting a number of documents, N−, in D− divided by the total number of documents, N, in D; obtain a value Ao as P++P−; obtain a probability, P+c, by counting a number of documents, N+c, in D+c divided by the number of documents, N, in D; obtain a probability, P−c, by counting a number of documents, N−c, in D−c divided by the number of documents, N, in D; obtain a probability, P+(c-1), by counting a number of documents, N+(c-1), in D−(c-1) divided by the number of documents, N, in D; obtain a probability, P−(c-1), by counting a number of documents, N−(c-1), in D−(c-1) divided by the number of documents, N, in D; obtain a value Ae as a probability of obtaining a positive document, P+c*P+(c-1), plus a probability of obtaining a negative document, P−c*P−(c-1); obtain a Kappa value L as Ao−Ae divided by (1−Ao); compare the Kappa value L to the stopping threshold tstop to determine whether tstop≦Ao−Ae/1−Ae which indicates that the stopping criteria has been met; and, based on the determination that the stopping criteria has not been met, store the score vector Sc to memory as a previous score vector Sc-1. 9. The apparatus of claim 1, wherein the apparatus is further operable to: discriminate, using the updated hyperplane h(x), remaining documents from the unlabeled set of available documents D where each of the remaining documents are classified as relevant or non-relevant. 10. A method in an apparatus for implementing batch-mode active learning for technology-assisted review (TAR) of documents, the method comprising: obtaining an unlabeled set of documents D; obtaining a batch size k; constructing a first batch of k documents D; obtaining labels for the first batch of k documents D, wherein the labeled first batch of k documents D are referred to as training data documents; constructing a classification model, Mc, using the labeled first batch of k documents D; performing an iteration of active learning using the classification model Mc, wherein the perform operation comprises: (i) selecting a new batch of unlabeled instances Bc using a current version of the classification model, Mc(x), an unlabeled set of available documents D, and the batch size k, wherein the apparatus is further operable to implement a Diversity Sampler process or a Biased Probabilistic Sampler process to select the new batch of unlabeled instances Bc; (ii) obtaining labels for the new batch of unlabeled instances Bc; (iii) adding the labeled new batch of instances Bc to a current version of the training data documents referred to as extended training data documents Dc; constructing an updated classification model M(x) using the extended training data documents Dc; determining whether a stopping criteria has been met; based on the determination that the stopping criteria has not been met, repeating the performing step, the third constructing step, and the determining step; and, based on the determination that the stopping criteria has been met, returning the updated classification model M(x). 11. The method of claim 10, wherein the classification model, Mc, comprises one of the following: a support vector machine model, a logistic regression model, a nearest neighbors model, decision forest model, neural network model, Bayesian model, or ensemble model 12. The method of claim 10, wherein the Diversity Sampler process comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), and a similarity threshold t; sorting the unlabeled set of available documents D using a sorting function to obtain sorted indices I of the unlabeled set of available documents D; inserting the sorted document having a nearest sorted indice I[1] of the sorted unlabeled set of available documents D into the new batch of unlabeled instances Bc; obtaining sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; removing the documents with the obtained sorted indices I from the sorted unlabeled set of available documents D; and repeating the inserting step, the second obtaining step, and the removing step until insert k documents into the new batch of unlabeled instances Bc. 13. The method of claim 12, wherein the sorting function is one of the following: a distance from a hyperplane based on smallest to greatest when the classification model Mc(x) is a support vector machine, SVM, classification model; a distance from a hyperplane based on greatest to smallest when the classification model Mc(x) is a support vector machine, SVM, classification model; or a random sorting function; an entropy function; a least confidence function; or a committee disagreement function. 14. The method of claim 10, wherein the Biased Probabilistic Sampler process comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), a similarity threshold t, and a probability distribution function P(x); constructing a probability vector w based on an probability distribution function P(x) for each document from the unlabeled set of available documents D; choosing a document I[1] from the unlabeled set of available documents D using the weight w of the probability vector; inserting the chosen document I[1] into the new batch of unlabeled instances Bc; obtaining sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; removing the documents with the obtained sorted indices I from the unlabeled set of available documents D; re-normalizing the weight vector as documents have been removed from the sorted unlabeled set of available documents D; and, repeating the choosing step, the inserting step, the second obtaining step, the removing step, and the re-normalizing step until select the new batch of unlabeled instances Bc. 15. The method of claim 10, wherein the step of determining whether the stopping criteria has been met further comprises implementing a Kappa agreement which is a measure of how much the current version of the hyperplane hc(x) and the updated hyperplane h(x) agree on respective prediction of labels of a chosen set of the documents D. 16. The method of claim 10, wherein the step of determining whether the stopping criteria has been met further comprises implementing an Accumulated History of Scores process which comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, an accumulation function A(Sc, Sc-1, . . . , S1), and a stopping threshold tstop; constructing a score vector Sc using the current classification model Mc(x) and the total set of documents DT; combining a current score vector Sc with previous score vectors (Sc-1, . . . , S1) using the accumulation function A(Sc, Sc-1, . . . , S1) to obtain a stability value (s); comparing the stability value s to the stopping threshold tstop to determine whether tstop≦S which indicates that the stopping criteria has been met; and based on the determination that the stopping criteria has not been met, storing the current score vector Sc to memory as a previous score vector Sc-1. 17. The method of claim 10, wherein the step of determining whether the stopping criteria has been met further comprises implementing a Cohen's Kappa process which comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, and a stopping threshold tstop; constructing a score vector Sc using the current classification model Mc(x) and the total set of documents DT; retrieving a previous score vector Sc-1; obtaining a set of documents D+c from the documents DT that have a positive score in Sc; obtaining a set of documents D−c from the documents DT that have a negative score in Sc obtaining a set of documents D+(c-1) from the documents DT that have a positive score in Sc-1; obtaining a set of documents D−(c-1) from the documents DT that have a negative score in Sc-1; obtaining a set of documents in common, D+, between D+c and D+(c-1); obtaining a set of documents in common, D−, between D−c and D−(c-1); obtaining a probability, P+, of a document having positive score in both score vectors Sc and Sc-1 by counting a number of documents, N+, in D+ divided by the total number of documents, N, in D; obtaining a probability, P−, of a document having negative score in both score vectors Sc and Sc-1, by counting a number of documents, N−, in D− divided by the total number of documents, N, in D; obtaining a value Ao as P++P−; obtaining a probability, P+c, by counting a number of documents, N+c, in D+c divided by the number of documents, N, in D; obtaining a probability, P−c, by counting a number of documents, N−c, in D−c divided by the number of documents, N, in D; obtaining a probability, P+(c-1), by counting a number of documents, N+(c-1), in D+(c-1) divided by the number of documents, N, in D; obtaining a probability, P−(c-1), by counting a number of documents, N−(c-1), in D−(c-1) divided by the number of documents, N, in D; obtaining a value Ae as a probability of obtaining a positive document, P+c*P+(c-1), plus a probability of obtaining a negative document, P−c*P−(c-1); obtaining a Kappa value L as Ao−Ae divided by (1−Ao); comparing the Kappa value L to the stopping threshold tstop to determine whether tstop≦Ao−Ae/1−Ae which indicates that the stopping criteria has been met; and, based on the determination that the stopping criteria has not been met, storing the score vector Sc to memory as a previous score vector Sc-1. 18. The method of claim 10, further comprising: discriminating, using the updated hyperplane h(x), remaining documents from the unlabeled set of available documents D where each of the remaining documents are classified as relevant or non-relevant. 19. An apparatus configured to implement batch-mode active learning for technology-assisted review (TAR) of documents, the apparatus comprising: a processor; and, a memory that stores processor-executable instructions, wherein the processor interfaces with the memory to execute the processor-executable instructions, whereby the apparatus is operable to: obtain an unlabeled set of documents D; obtain a batch size k; construct a first batch of k documents D; obtain labels for the first batch of k documents D, wherein the labeled first batch of k documents D are referred to as training data documents; construct a classification model, Mc, using the labeled first batch of k documents D; perform an iteration of active learning using the classification model Mc, wherein the perform operation comprises: (i) select a new batch of unlabeled instances Bc using a current version of the classification model, Mc(x), an unlabeled set of available documents D, and the batch size k; (ii) obtain labels for the new batch of unlabeled instances Bc; (iii) add the labeled new batch of instances Bc to a current version of the training data documents referred to as extended training data documents Dc; construct an updated classification model M(x) using the extended training data documents Dc; determine whether a stopping criteria has been met, wherein the apparatus is further operable to determine whether the stopping criteria has been met by implementing an Accumulated History of Scores process which comprises the following operations: (i) obtain the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, an accumulation function A(Sc, Sc-1, . . . , S1), and a stopping threshold tstop; (ii) construct a score vector Sc using the current classification model Mc(x) and the total set of documents DT; (iii) combine a current score vector Sc with previous score vectors (Sc-1, . . . , S1) using the accumulation function A(Sc, Sc-1, . . . , S1) to obtain a stability value (s); (iv) compare the stability value s to the stopping threshold tstop to determine whether tstop≦S which indicates that the stopping criteria has been met; and (v) based on the determination that the stopping criteria has not been met, store the current score vector Sc to memory as a previous score vector Sc-1 based on the determination that the stopping criteria has not been met, repeat the perform operation, the third construct operation, and the determine operation; and, based on the determination that the stopping criteria has been met, return the updated classification model M(x). 20. The apparatus of claim 19, wherein the classification model, Mc, comprises one of the following: a support vector machine model, a logistic regression model, a nearest neighbors model, decision forest model, neural network model, Bayesian model, or ensemble model. 21. The apparatus of claim 19, wherein the apparatus is further operable to implement a Diversity Sampler process to select the new batch of unlabeled instances Bc as follows: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), and a similarity threshold t; sort the unlabeled set of available documents D using a sorting function to obtain sorted indices I of the unlabeled set of available documents D; insert the sorted document having a nearest sorted indice I[1] of the sorted unlabeled set of available documents D into the new batch of unlabeled instances Bc; obtain sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; remove the documents with the obtained sorted indices I from the sorted unlabeled set of available documents D; and repeat the insert operation, the second obtain operation, and the remove operation until insert k documents into the new batch of unlabeled instances Bc. 22. The apparatus of claim 21, wherein the sorting function is one of the following: a distance from a hyperplane based on smallest to greatest when the classification model Mc(x) is a support vector machine, SVM, classification model; a distance from a hyperplane based on greatest to smallest when the classification model Mc(x) is a support vector machine, SVM, classification model; or a random sorting function; an entropy function; or a least confidence function; or a committee disagreement function. 23. The apparatus of claim 19, wherein the apparatus is further operable to implement a Biased Probabilistic Sampler process to select the new batch of unlabeled instances Bc as follows: obtain the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), a similarity threshold t, and a probability distribution function P(x); construct a probability vector w based on an probability distribution function P(x) for each document from the unlabeled set of available documents D; choose a document I[1] from the unlabeled set of available documents D using the weight w of the probability vector; insert the chosen document I[1] into the new batch of unlabeled instances Bc; obtain sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; remove the documents with the obtained sorted indices I from the unlabeled set of available documents D; re-normalize the weight vector as documents have been removed from the sorted unlabeled set of available documents D; and, repeat the choose operation, the insert operation, the second obtain operation, the remove operation, and the re-normalize operation until select the new batch of unlabeled instances Bc. 24. The apparatus of claim 19, wherein the apparatus is further operable to: discriminate, using the updated hyperplane h(x), remaining documents from the unlabeled set of available documents D where each of the remaining documents are classified as relevant or non-relevant. 25. A method in an apparatus for implementing batch-mode active learning for technology-assisted review (TAR) of documents, the method comprising: obtaining an unlabeled set of documents D; obtaining a batch size k; constructing a first batch of k documents D; obtaining labels for the first batch of k documents D, wherein the labeled first batch of k documents D are referred to as training data documents; constructing a classification model, Mc, using the labeled first batch of k documents D; performing an iteration of active learning using the classification model Mc, wherein the perform operation comprises: (i) selecting a new batch of unlabeled instances Bc using a current version of the classification model, Mc(x), an unlabeled set of available documents D, and the batch size k; (ii) obtaining labels for the new batch of unlabeled instances Bc; (iii) adding the labeled new batch of instances Bc to a current version of the training data documents referred to as extended training data documents Dc; constructing an updated classification model M(x) using the extended training data documents Dc; determining whether a stopping criteria has been met, wherein the apparatus is further operable to determine whether the stopping criteria has been met by implementing an Accumulated History of Scores process which comprises the following steps: (i) obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D and the extended training data documents Dc referred to as the total set of documents DT, an accumulation function A(Sc, Sc-1, . . . , S1), and a stopping threshold tstop; (ii) constructing a score vector Sc using the current classification model Mc(x) and the total set of documents DT; (iii) combining a current score vector Sc with previous score vectors (Sc-1, . . . , S1) using the accumulation function A(Sc, Sc-1, . . . , S1) to obtain a stability value (s); (iv) comparing the stability value s to the stopping threshold tstop to determine whether tstop≦S which indicates that the stopping criteria has been met; and (v) based on the determination that the stopping criteria has not been met, storing the current score vector Sc to memory as a previous score vector Sc-1 based on the determination that the stopping criteria has not been met, repeating the performing step, the third constructing step, and the determining step; and, based on the determination that the stopping criteria has been met, returning the updated classification model M(x). 26. The method of claim 25, wherein the classification model, Mc, comprises one of the following: a support vector machine model, a logistic regression model, a nearest neighbors model, decision forest model, neural network model, Bayesian model, or ensemble model. 27. The method of claim 25, wherein the step of selecting the new batch of unlabeled instances Bc further comprising implementing a Diversity Sampler process which comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), and a similarity threshold t; sorting the unlabeled set of available documents D using a sorting function to obtain sorted indices I of the unlabeled set of available documents D; inserting the sorted document having a nearest sorted indice I[1] of the sorted unlabeled set of available documents D into the new batch of unlabeled instances Bc; obtaining sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; removing the documents with the obtained sorted indices I from the sorted unlabeled set of available documents D; and repeating the inserting step, the second obtaining step, and the removing step until insert k documents into the new batch of unlabeled instances Bc. 28. The apparatus of claim 27, wherein the sorting function is one of the following: a distance from a hyperplane based on smallest to greatest when the classification model Mc(x) is a support vector machine, SVM, classification model; a distance from a hyperplane based on greatest to smallest when the classification model Mc(x) is a support vector machine, SVM, classification model; or a random sorting function; an entropy function; a least confidence function; or a committee disagreement function. 29. The method of claim 25, wherein the step of selecting the new batch of unlabeled instances Bc further comprising implementing a Biased Probabilistic Sampler process which comprises the following steps: obtaining the current version of the classification model Mc(x), the unlabeled set of available documents D, the batch size k, a similarity function S(x1, x2), a similarity threshold t, and a probability distribution function P(x); constructing a probability vector w based on an probability distribution function P(x) for each document from the unlabeled set of available documents D; choosing a document I[1] from the unlabeled set of available documents D using the weight w of the probability vector; inserting the chosen document I[1] into the new batch of unlabeled instances Bc; obtaining sorted indices I of the sorted unlabeled set of available documents D that have a similarity score S(XI[1], X)≧t with respect to the inserted document I[1]; removing the documents with the obtained sorted indices I from the unlabeled set of available documents D; re-normalizing the weight vector as documents have been removed from the sorted unlabeled set of available documents D; and, repeating the choosing step, the inserting step, the second obtaining step, the removing step, and the re-normalizing step until select the new batch of unlabeled instances Bc. 30. The method of claim 25, further comprising: discriminating, using the updated hyperplane h(x), remaining documents from the unlabeled set of available documents D where each of the remaining documents are classified as relevant or non-relevant. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: The present disclosure relates to the electronic document review field and, more particularly, to various apparatuses and methods of implementing batch-mode active learning for technology-assisted review (TAR) of documents (e.g., legal documents). |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The present disclosure relates to the electronic document review field and, more particularly, to various apparatuses and methods of implementing batch-mode active learning for technology-assisted review (TAR) of documents (e.g., legal documents). |
|
A method, system and computer-usable medium are disclosed for adjusting fact-based answers provided by a question/answer (QA) system. A user submits a question to the QA system, where it is categorized into a question type. The QA system then processes the question to generate an answer. The QA system then generates an answer adjustment if it is determined that the question type and answer meet a predicted undesirable outcome. The answer adjustment may include a warning, a disclaimer, a recommendation, an alternative fact-based answer, a referral to an assistance service, or any combination thereof. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method for adjusting answers to a question asked of a question/answer (QA) system, comprising: receiving a question as user input from a user, the question being received by the QA system; categorizing the question into a question type; processing the question to generate an answer; and generating an answer adjustment if it is determined that the question type and the answer correspond to a potentially undesirable outcome criteria. 2. The method of claim 1, wherein the answer adjustment is at least one member of the set of: a warning; a disclaimer; an alternative answer an assistance service; a recommendation; and a referral to an assistance service. 3. The method of claim 1, wherein the QA system is trained to identify a question that might lead to an undesirable outcome. 4. The method of claim 3, wherein the method of the training is selected from a group consisting of: manual; automated; and learning. 5. The method of claim 1, wherein the answer and the answer adjustment are provided to the user. 6. The method of claim 1, further comprising generating and retaining a time-stamped record comprising: the question; the answer; the answer adjustment; and identification information associated with the user. 7-20. (canceled) |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method, system and computer-usable medium are disclosed for adjusting fact-based answers provided by a question/answer (QA) system. A user submits a question to the QA system, where it is categorized into a question type. The QA system then processes the question to generate an answer. The QA system then generates an answer adjustment if it is determined that the question type and answer meet a predicted undesirable outcome. The answer adjustment may include a warning, a disclaimer, a recommendation, an alternative fact-based answer, a referral to an assistance service, or any combination thereof. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method, system and computer-usable medium are disclosed for adjusting fact-based answers provided by a question/answer (QA) system. A user submits a question to the QA system, where it is categorized into a question type. The QA system then processes the question to generate an answer. The QA system then generates an answer adjustment if it is determined that the question type and answer meet a predicted undesirable outcome. The answer adjustment may include a warning, a disclaimer, a recommendation, an alternative fact-based answer, a referral to an assistance service, or any combination thereof. |
|
Embodiments of the present invention provide an intelligent interaction method and an intelligent interaction system which are directed to resolve the problem that the conventional intelligent interaction methods are too simple and the interaction effect is not good since the response information is based on the acquired intention information only. The intelligent interaction method comprises: acquiring current request information from a user and user static information corresponding to the user; performing intention analysis on the current request information to acquire intention information corresponding to the current request information; acquiring interaction background information corresponding to the user static information; and acquiring response information according to the intention information and the interaction background information and sending the response information to the user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An intelligent interaction method, comprising: acquiring current request information from a user and user static information corresponding to the user; performing intention analysis on the current request information to acquire intention information corresponding to the current request information; acquiring interaction background information corresponding to the user static information; and acquiring response information according to the intention information and the interaction background information and sending the response information to the user. 2. The method of claim 1, wherein acquiring response information according to the intention information and the interaction background information comprises: acquiring a corresponding response identifier according to the intention information and the interaction background information; and acquiring the response information according to the intention information and the response identifier, wherein correspondence relationship among the intention information, the interaction background information and the response identifier is pre-established, and correspondence relationship among the intention information, the response identifier and the response information is pre-established. 3. The method of claim 1, wherein acquiring response information according to the intention information and the interaction background information comprises: acquiring a corresponding response identifier according to the intention information and the interaction background information; and acquiring the response information according to the response identifier, wherein correspondence relationship among the intention information, the interaction background information and the response identifier is pre-established, and correspondence relationship between the response identifier and the response information is pre-established. 4. The method of claim 1, wherein acquiring interaction background information corresponding to the user static information comprises: classifying the user static information, and acquiring the interaction background information corresponding to the user static information based on a classification result. 5. The method of claim 4, wherein the interaction background information comprises at least one interaction background item, each interaction background item comprises at least one type of interaction background contents. 6. The method of claim 5, wherein classifying the user static information and acquiring the interaction background information corresponding to the user static information based on a classification result comprises: classifying the user static information into at least one static information category; and determining all types of interaction background contents matching the user static information according to the static information categories included in the user static information, each type of the interaction background contents being determined according to at least one of the static information categories. 7. An intelligent interaction system, comprising a knowledge base, an interaction module, an intention analysis module, a background acquisition module and a response decision module, wherein the knowledge base is configured to store intention information, interaction background information and response information; the interaction module is configured to acquire current request information from a user and user static information corresponding to the user, and send response information acquired by the response decision module to the user; the intention analysis module is configured to perform intention analysis on the current request information acquired by the interaction module to acquire the intention information corresponding to the current request information from the knowledge base; the background acquisition module is configured to acquire the interaction background information corresponding to the user static information from the knowledge base; and the response decision module is configured to acquire the response information from the knowledge base according to the intention information and the interaction background information. 8. The system of claim 7, wherein the knowledge base is further configured to store pre-established correspondence relationship among the intention information, the interaction background information and response identifier, and pre-established corresponding relationship among the intention information, the response identifier and the response information; wherein the response decision module comprises: a response identifier acquisition unit configured to acquire a corresponding response identifier from the knowledge base according to the intention information and the interaction background information; and a control unit configured to acquire the response information from the knowledge base according to the intention information and the response identifier. 9. The system of claim 7, wherein the knowledge base is further configured to store pre-established correspondence relationship among the intention information, the interaction background information and response identifier, and pre-established corresponding relationship between the response identifier and the response information; wherein the response decision module comprises: a response identifier acquisition unit configured to acquire a corresponding response identifier from the knowledge base according to the intention information and the interaction background information; and a control unit configured to acquire the response information from the knowledge base according to the response identifier. 10. The system of claim 8, wherein the response identifier is any of a response tone identifier which is classified into at least two categories from a gentle tone to a stern tone, a response pitch identifier that is classified into at least two categories from a high pitch to a low pitch, a response speed identifier which is classified into at least two categories from a low speed to a high speed, and a response volume identifier which is classified into at least two categories from a low volume to a high volume, or any combination thereof. 11. The system of claim 8, wherein the correspondence relationship among the intention information, the interaction background information and the response identifier and the correspondence relationship among the intention information, the response identifier and the response information is pre-established through large data classification and clustering technology. 12. The system of claim 7, wherein the background acquisition module is further configured to classify the user static information and acquire the interaction background information corresponding to the user static information from the knowledge base based on a classification result. 13. The system of claim 12, wherein the interaction background information comprises at least one interaction background item, each interaction background item comprises at least one type of interaction background contents. 14. The system of claim 13, wherein the background acquisition module comprises: a classification unit configured to classify the user static information into at least one static information category; and an acquisition unit configured to determine all interaction background contents matching the user static information according to the static information categories included in the user static information, each of the interaction background contents being determined according to at least one of the static information categories. 15. The system of claim 14, wherein the user static information is classified into at least one of the following static information categories: credit card service attribute, user identity information, credit card type, current billing period, total debt and the amount has been repaid; wherein the interaction background information includes at least one of the following interaction background items: credit card service attribute, current billing period, debt history status and repayment history status; wherein the credit card service attribute includes at least one of the following interaction background contents: credit card debt, new credit card application, credit line inquiry, credit card repayment and credit card cancellation; wherein the debt history status includes two interaction background contents: never and ever; and wherein the repayment history status includes at least one of the following interaction background contents: no repayment, partial repayment and interval repayment. 16. The system of claim 7, wherein the interaction module is further configured to acquire the user static information through user input or interaction with a third party. 17. The system of claim 7, wherein the intention analysis module comprises: a matching unit configured to match text contents of the current request information with a plurality of preset semantic templates to determine a matching semantic template; and an intention determination unit configured to acquire the intention information corresponding to the matching semantic template, wherein corresponding relationship between the semantic templates and the intention information is pre-established and stored in the knowledge base, same intention information corresponds to one or more of the semantic templates. 18. The system of claim 17, wherein the matching unit is further configured to calculate similarity between the text contents of the current request information and the plurality of preset semantic templates, and select one semantic template having the highest similarity as the matching semantic template. 19. The system of claim 17, wherein the current request information is a voice message, the interaction module further comprises a text conversion unit configured to convert the current request information into a text message. 20. The system of claim 7, wherein the interaction module further comprises a voice conversion unit configured to convert the response information into a voice message to be sent to the user. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Embodiments of the present invention provide an intelligent interaction method and an intelligent interaction system which are directed to resolve the problem that the conventional intelligent interaction methods are too simple and the interaction effect is not good since the response information is based on the acquired intention information only. The intelligent interaction method comprises: acquiring current request information from a user and user static information corresponding to the user; performing intention analysis on the current request information to acquire intention information corresponding to the current request information; acquiring interaction background information corresponding to the user static information; and acquiring response information according to the intention information and the interaction background information and sending the response information to the user. |
|
G06N3006 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Embodiments of the present invention provide an intelligent interaction method and an intelligent interaction system which are directed to resolve the problem that the conventional intelligent interaction methods are too simple and the interaction effect is not good since the response information is based on the acquired intention information only. The intelligent interaction method comprises: acquiring current request information from a user and user static information corresponding to the user; performing intention analysis on the current request information to acquire intention information corresponding to the current request information; acquiring interaction background information corresponding to the user static information; and acquiring response information according to the intention information and the interaction background information and sending the response information to the user. |
|
The present disclosure provides systems and methods that enable training of an encoder model based on a decoder model that performs an inverse transformation relative to the encoder model. In one example, an encoder model can receive a first set of inputs and output a first set of outputs. The encoder model can be a neural network. The decoder model can receive the first set of outputs and output a second set of outputs. A loss function can describe a difference between the first set of inputs and the second set of outputs. According to an aspect of the present disclosure, the loss function can be sequentially backpropagated through the decoder model without modifying the decoder model and then through the encoder model while modifying the encoder model, thereby training the encoder model. Thus, an encoder model can be trained to have enforced consistency relative to the inverse decoder model. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method to perform machine learning, the method comprising: obtaining, by one or more computing devices, data descriptive of an encoder model that is configured to receive a first set of inputs and, in response to receipt of the first set of inputs, output a first set of outputs; obtaining, by the one or more computing devices, data descriptive of a decoder model that is configured to receive the first set of outputs and, in response to receipt of the first set of outputs, output a second set of outputs; determining, by the one or more computing devices, a loss function that describes a difference between the first set of inputs and the second set of outputs; backpropagating, by the one or more computing devices, the loss function through the decoder model without modifying the decoder model; and after backpropagating, by the one or more computing devices, the loss function through the decoder model, continuing to backpropagate, by the one or more computing devices, the loss function through the encoder model to train the encoder model; wherein continuing to backpropagate, by the one or more computing devices, the loss function through the encoder model to train the encoder model comprises adjusting, by the one or more computing devices, at least one weight included in the encoder model. 2. The computer-implemented method of claim 1, wherein: the encoder model is configured to: receive the first set of inputs expressed according to a first set of dimensions; and output the first set of outputs expressed according to a second set of dimensions that are different from the first set of dimensions; and the decoder model is configured to output the second set of outputs expressed according to the first set of dimensions. 3. The computer-implemented method of claim 1, wherein at least the encoder model comprises a neural network. 4. The computer-implemented method of claim 1, wherein the encoder model comprises a sensor fusion model that is configured to: receive a set of sensor data as the first set of inputs, the set of sensor data reported by a plurality of sensors; and in response to receipt of the set of sensor data, output a set of condition data as the first set of outputs, the set of condition data descriptive of a condition evidenced by the set of sensor data. 5. The computer-implemented method of claim 4, wherein the decoder model comprises a sensor data prediction model that is configured to: receive the set of condition data; and in response to receipt of the set of condition data, predict a second set of sensor data, the second set of sensor data comprising sensor readings expected to result from the condition described by the set of condition data. 6. The computer-implemented method of claim 1, wherein the encoder model comprises a sensor fusion model that is configured to: receive a set of sensor data as the first set of inputs, the set of sensor data reported by a plurality of sensors of a mobile device; and in response to receipt of the set of sensor data, output a set of pose data as the first set of outputs, the set of pose data descriptive of a pose of the mobile device that includes the plurality of sensors. 7. The computer-implemented method of claim 6, wherein the set of pose data comprises a set of six degree of freedom pose data that describes the pose of the mobile device in six degrees of freedom. 8. The computer-implemented method of claim 6, wherein the decoder model comprises a sensor prediction model that is configured to: receive the set of pose data; and in response to receipt of the set of pose data, predict a second set of sensor data, the second set of sensor data comprising sensor readings expected to result from the pose of the mobile device. 9. The computer-implemented method of claim 8, further comprising: determining, by the one or more computing devices, the pose of the mobile device based at least in part on the set of pose data output by the sensor fusion model. 10. A computing system to perform machine learning, the computing system comprising: at least one processor; and at least one tangible, non-transitory computer-readable medium that stores instructions that, when executed by the at least one processor, cause the computing system to: obtain data descriptive of a model that comprises an encoder model and a decoder model, wherein the encoder model is configured to receive a first set of inputs and, in response to receipt of the first set of inputs, output a first set of outputs, and wherein the decoder model is configured to receive the first set of outputs and, in response to receipt of the first set of outputs, output a second set of outputs; determine a loss function that describes a difference between the first set of inputs and the second set of outputs; backpropagate the loss function through the decoder model without modifying the decoder model; and after backpropagating the loss function through the decoder model, continue to backpropagate the loss function through the encoder model while modifying the encoder model to train the encoder model. 11. The computing system of claim 10, wherein: the encoder model is configured to receive the first set of inputs expressed according to a first set of dimensions; the encoder model is configured to output the first set of outputs expressed according to a second set of dimensions in response to receipt of the first set of inputs; and the decoder model is configured to output the second set of outputs expressed according to the first set of dimensions. 12. The computing system of claim 10, wherein at least the encoder model comprises a neural network. 13. The computing system of claim 10, wherein: the encoder model comprises a sensor fusion model that is configured to: receive a set of sensor data as the first set of inputs, the set of sensor data reported by a plurality of sensors of a mobile device; and in response to receipt of the set of sensor data, output a set of pose data as the first set of outputs, the set of pose data descriptive of a pose of the mobile device that includes the plurality of sensors; and the decoder model comprises a sensor prediction model that is configured to: receive the set of pose data; and in response to receipt of the set of pose data, predict a second set of sensor data, the second set of sensor data comprising sensor readings expected to result from the pose of the mobile device. 14. The computing system of claim 13, wherein: the computing system consists of the mobile device; the mobile device comprises the plurality of sensors, the at least one processor, and the at least one tangible, non-transitory computer-readable medium that stores the instructions; and the at least one tangible, non-transitory computer-readable medium stores the model. 15. The computing system of claim 14, wherein execution of the instructions by the at least one processor further causes the mobile device to, after continuing to backpropagate the loss function through the sensor fusion model to train the sensor fusion model: receive a third set of sensor data newly reported by the plurality of sensors; input the third set of sensor data into the sensor fusion model; receive a second set of pose data as an output of the sensor fusion model; and determine a current pose of the mobile device based at least in part on the second set of pose data. 16. The computing system of claim 10, wherein: the encoder model comprises a computer vision model that is configured to: receive a set of image data as the first set of inputs, the set of image data descriptive of one or more first frames of imagery that depict a scene; and in response to receipt of the set of image data, output a set of depth data as the first set of outputs, the set of depth data descriptive of one or more depths associated with the scene depicted by the one or more frames of imagery; and the decoder model comprises an image rendering model that is configured to: receive the set of depth data; and in response to receipt of the set of depth data, predict a second set of image data, the second set of image data comprising one or more second frames of imagery that depict the expected appearance of the scene in view of the set of depth data. 17. The computing system of claim 10, wherein: the encoder model comprises a speech-to-text model that is configured to: receive a set of audio data as the first set of inputs, the set of audio data descriptive of an utterance; and in response to receipt of the set of audio data, output a set of textual data as the first set of outputs, the set of textual data providing a textual transcript of the utterance; and the decoder model comprises a text-to-speech model that is configured to: receive the set of textual data; and in response to receipt of the set of textual data, predict a second set of audio data, the second set of audio data comprising a recreated utterance of the textual transcript. 18. A computing system, comprising: at least one processor; and at least one memory that stores a machine-learned encoder model that is configured to receive a first set of inputs and output a first set of outputs, the encoder model having been trained by sequentially backpropagating a loss function through a decoder model without modifying the decoder model and then through the encoder model to modify at least one weight of the encoder model, the decoder model configured to receive the first set of outputs and output a second set of outputs, the loss function descriptive of a difference between the first set of inputs and the second set of outputs. 19. The computing system of claim 18, wherein the encoder model comprises a neural network. 20. The computing system of claim 18, wherein: the encoder model comprises a sensor fusion model; and the computing system is configured to: receive a set of sensor data reported by a plurality of sensors of a mobile device; input the set of sensor data into the sensor fusion model; and receive a set of pose data as an output of the encoder model. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: The present disclosure provides systems and methods that enable training of an encoder model based on a decoder model that performs an inverse transformation relative to the encoder model. In one example, an encoder model can receive a first set of inputs and output a first set of outputs. The encoder model can be a neural network. The decoder model can receive the first set of outputs and output a second set of outputs. A loss function can describe a difference between the first set of inputs and the second set of outputs. According to an aspect of the present disclosure, the loss function can be sequentially backpropagated through the decoder model without modifying the decoder model and then through the encoder model while modifying the encoder model, thereby training the encoder model. Thus, an encoder model can be trained to have enforced consistency relative to the inverse decoder model. |
|
G06N3084 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The present disclosure provides systems and methods that enable training of an encoder model based on a decoder model that performs an inverse transformation relative to the encoder model. In one example, an encoder model can receive a first set of inputs and output a first set of outputs. The encoder model can be a neural network. The decoder model can receive the first set of outputs and output a second set of outputs. A loss function can describe a difference between the first set of inputs and the second set of outputs. According to an aspect of the present disclosure, the loss function can be sequentially backpropagated through the decoder model without modifying the decoder model and then through the encoder model while modifying the encoder model, thereby training the encoder model. Thus, an encoder model can be trained to have enforced consistency relative to the inverse decoder model. |
|
Various implementations for assigning rules and creating rules using templates are described herein. In one example implementation, a model is determined, one or more components of the model are determined, a rule from a set of one or more predefined rules is determined, and the rule is assigned to the model. The rule has one or more parameters matching the one or more components of the model. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. (canceled) 2. A computer-implemented method comprising: receiving input defining a set of global rules for configuring models, each of the set of global rules including one or more rule conditions that trigger a rule and one or more rule actions associated with one or more attributes, the one or more rule actions being performed when a rule is triggered; after defining the set of global rules, receiving input to configure a model; determining one or more components of the model; determining a rule from the set of global rules having one or more of the one or more rule conditions and the one or more rule actions matching one or more of the one or more components of the model; and assigning the rule to the model by one of an explicit assignment and an implicit assignment. 3. The computer-implemented method of claim 2, wherein determining the rule from the set of global rules includes: matching the one or more components of the model to one or more attributes, option groups, and products; and determining the rule based on the matching. 4. The computer-implemented method of claim 3, wherein the rule includes the one or more attributes, option groups, and products, and the one or more attributes, option groups, and products correspond to the one or more components of the model. 5. The computer-implemented method of claim 3, further comprising: overriding the rule that is assigned to the model by assigning the rule differently to the model or assigning another rule to the model. 6. The computer-implemented method of claim 5, further comprising: providing a rule definition interface including the assignment of the rule to the model for presentation to a user; and receiving an input from the user to override the assignment of the rule. 7. The computer-implemented method of claim 2, wherein determining the rule from the set of global rules includes automatically determining the rule based on a common attribute between the rule and the model. 8. The computer-implemented method of claim 7, wherein the common attribute includes an option group, attribute, or product, and the model and the rule each includes the option group, attribute, or product. 9. The computer-implemented method of claim 2, further comprising: determining an option group associated with the model; determining that the option group includes a product; and determining that the rule is dependent on the product, wherein the rule is assigned to the model based on the rule being dependent on the product. 10. The computer-implemented method of claim 2, further comprising: determining an option group associated with the model; determining that the option group is associated with an attribute; and determining that the rule is dependent on the attribute, wherein the rule is assigned to the model based on the rule being dependent on the attribute. 11. The computer-implemented method of claim 2, further comprising: determining an option group associated with the model is user-defined; determining an attribute tied to an option of the option group; and determining a rule is dependent on the attribute tied to the option of the option group, wherein the rule is assigned to the model based on the rule being dependent on the attribute. 12. A system comprising: one or more processors; one or more memories; a modeling engine embodied by instructions stored in the one or more memories and, when executed by the one or more processors, perform operations comprising: receiving input defining a set of global rules for configuring models, each of the set of global rules including one or more rule conditions that trigger a rule and one or more rule actions associated with one or more attributes, the one or more rule actions being performed when a rule is triggered; after defining the set of global rules, receiving input to configure a model; determining one or more components of the model; determining a rule from the set of global rules having one or more of the one or more rule conditions and the one or more rule actions matching one or more of the one or more components of the model; and assigning the rule to the model by one of an explicit assignment and an implicit assignment. 13. The system of claim 12, wherein determining the rule from the set of global rules includes: matching the one or more constituent components of the model to one or more attributes, option groups, and products; and determining the rule based on the matching. 14. The system of claim 13, wherein the rule includes the one or more attributes, option groups, and products, and the one or more attributes, option groups, and products correspond to the one or more constituent components of the model. 15. The system of claim 13, wherein the operations further comprise: overriding the rule that is assigned to the model by assigning the rule differently to the model or assigning another rule to the model. 16. The system of claim 15, wherein the operations further comprise: providing a rule definition interface including the assignment of the rule to the model for presentation to a user; and receiving an input from the user to override the assignment of the rule. 17. The system of claim 12, wherein determining the rule from the set of global rules includes automatically determining the rule based on a common attribute between the rule and the model. 18. The system of claim 17, wherein the common attribute includes an option group, attribute, or product, and the model and the rule each includes the option group, attribute, or product. 19. The system of claim 12, wherein the operations further comprise: determining an option group associated with the model; determining that the option group includes a product; and determining that the rule is dependent on the product, wherein the rule is assigned to the model based on the rule being dependent on the product. 20. The system of claim 12, wherein the operations further comprise: determining an option group associated with the model; determining that the option group is associated with an attribute; and determining that the rule is dependent on the attribute, wherein the rule is assigned to the model based on the rule being dependent on the attribute. 21. A system comprising: one or more processors; one or more memories storing instructions that, when executed by the one or more processors, perform operations comprising: receiving input defining a set of global rules for configuring models, each of the set of global rules including one or more rule conditions that trigger a rule and one or more rule actions associated with one or more attributes, the one or more rule actions being performed when a rule is triggered; after defining the set of global rules, receiving input to configure a model; determining one or more components of the model; determining a rule from the set of global rules having one or more of the one or more rule conditions and the one or more rule actions matching one or more of the one or more components of the model; and assigning the rule to the model by one of an explicit assignment and an implicit assignment. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Various implementations for assigning rules and creating rules using templates are described herein. In one example implementation, a model is determined, one or more components of the model are determined, a rule from a set of one or more predefined rules is determined, and the rule is assigned to the model. The rule has one or more parameters matching the one or more components of the model. |
|
G06N5025 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Various implementations for assigning rules and creating rules using templates are described herein. In one example implementation, a model is determined, one or more components of the model are determined, a rule from a set of one or more predefined rules is determined, and the rule is assigned to the model. The rule has one or more parameters matching the one or more components of the model. |
|
Methods are described for identifying events that would be considered surprising by people and identifying how and when to transmit information to a user about situations that they would likely find surprising. Additionally, the methods of identifying surprising situations can be used to build a case library of surprising events, joined with a set of observations before the surprising events occurred. Statistical machine learning methods can be applied with data from the case library to build models that can predict when a user will likely be surprised at future times. One or more models of context-sensitive expectations of people, a view of the current world, and methods for recording streams or events before surprises occur, and for building predictive models from a case library of surprises and such historical observations can be employed. The models of current and future surprises can be coupled with display and alerting machinery. | Please help me write a proper abstract based on the patent claims. CLAIM: 1-20. (canceled) 21. A method of providing user notifications regarding events, the method comprising: receiving a first predictive model, wherein the first predictive model is based on a user expectancy model designed to reflect whether a user would predict the event; receiving contextual data that is at least potentially associated with an event; with at least one processor, applying the first predictive model to at least a portion of the contextual data to generate a first prediction for the event; determining that the first prediction differs from a second prediction, the second prediction being based on a second predictive model that is different than the first predictive model; and in response to the determination, selectively outputting an indication regarding the event. 22. The method of claim 21, wherein: the first prediction is also based at least on a current time and/or day; the first prediction and the second prediction are different; and the indication represents that the event is atypical for the current time and/or day. 23. The method of claim 21, wherein: the first prediction is a prediction of a roadway traffic congestion condition and is also based at least on typical traffic congestion conditions for a current time of a day; the first prediction and the second prediction are different; and the indication represents that the event is atypical for the current time of the day. 24. The method of claim 23, wherein the indication also represents a prediction of atypically less than typical traffic congestion along a route. 25. The method of claim 21, wherein the second predictive model includes one or more statistical models that employ Bayesian networks, dynamic Bayesian networks, continuous time Bayesian networks, support vector machines, neural network models, Hidden Markov Models, Markov decision processes, and/or particle filtering. 26. The method of claim 21, wherein the first prediction is a prediction of roadway traffic conditions at a future time. 27. The method of claim 21, further comprising: calculating a measure of utility associated with outputting an indication for the predicted event; and outputting the indication if the measure of utility associated therewith is above a threshold. 28. The method of claim 21, wherein: the first predictive model also corresponds to human modes of predicting events; the second predictive model is expected to be more accurate than the first predictive model; and the indication is representative of a likelihood that the user would be surprised by the event. 29. A computer-readable storage device having instructions stored therein, the instructions, when executed on a computing device, cause the computing device to perform operations, the operations comprising: receiving contextual data that is at least potentially associated with an event; generating a first prediction for the event based on the received contextual data and on a user expectancy model, wherein the user expectancy model emulates predictive capabilities of a user; determining that the first prediction is different than a second prediction, wherein the second prediction is based on a second model that is different from the user expectancy model; and in response to the determination, selectively outputting an indication regarding the event. 30. The computer-readable storage device of claim 29, wherein: the second model is expected to be more accurate than the user expectancy model. 31. The computer-readable storage device of claim 29, wherein: the event is a traffic congestion event on a commute route of the user. 32. The computer-readable storage device of claim 29, wherein: the first prediction is a prediction of traffic congestion along a route and is based at least on typical traffic congestion along the route at a time of day. 33. The computer-readable storage device of claim 29, wherein: the indication represents a prediction of atypically less than typical traffic congestion along a route for a time of day. 34. The computer-readable storage device of claim 29, wherein: the user expectancy model corresponds to human modes of predicting events; the second model is expected to be more accurate than the user expectancy model; and the indication is representative of a likelihood that the user would be surprised by the event. 35. A computing device for providing notifications of events, the computing device comprising: a memory and a processor that respectively store and execute computer-executable instructions, including instructions that enable the computing device to: receive contextual data that is at least potentially associated with an event; generate a first prediction for the event based on the contextual data and on a user expectancy model, wherein the user expectancy model reflects human predictive capabilities; determine that the first prediction differs from a second prediction, the second prediction being based on a second predictive model that is different than the first predictive model; and in response to the determination, provide an indication regarding the event. 36. The computing device of claim 35, wherein the instructions further enable the computing device to selectively provide the indication based on an expectation that the indication would surprise a user of the computing device. 37. The computing device of claim 35, wherein the user expectancy model is based on historical data for other events similar to the event. 38. The computing device of claim 35, wherein: the first prediction is a prediction of a roadway traffic congestion condition and is based at least on typical traffic congestion conditions for a current time of day; the first prediction and the second prediction are different; and the indication represents that the event is atypical for the current time of day. 39. The computing device of claim 35, wherein the indication represents a prediction that a traffic congestion condition will happen at a future time on a commute route of a user of the computing device. 40. The computing device of claim 35, wherein: the second predictive model is expected to be more accurate than the user expectancy model. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods are described for identifying events that would be considered surprising by people and identifying how and when to transmit information to a user about situations that they would likely find surprising. Additionally, the methods of identifying surprising situations can be used to build a case library of surprising events, joined with a set of observations before the surprising events occurred. Statistical machine learning methods can be applied with data from the case library to build models that can predict when a user will likely be surprised at future times. One or more models of context-sensitive expectations of people, a view of the current world, and methods for recording streams or events before surprises occur, and for building predictive models from a case library of surprises and such historical observations can be employed. The models of current and future surprises can be coupled with display and alerting machinery. |
|
G06N7005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Methods are described for identifying events that would be considered surprising by people and identifying how and when to transmit information to a user about situations that they would likely find surprising. Additionally, the methods of identifying surprising situations can be used to build a case library of surprising events, joined with a set of observations before the surprising events occurred. Statistical machine learning methods can be applied with data from the case library to build models that can predict when a user will likely be surprised at future times. One or more models of context-sensitive expectations of people, a view of the current world, and methods for recording streams or events before surprises occur, and for building predictive models from a case library of surprises and such historical observations can be employed. The models of current and future surprises can be coupled with display and alerting machinery. |
|
Some embodiments are associated with a support vector machine having model parameters. According to some embodiments, a set of evaluation data may be received and a computer processor may automatically tune the model parameters during a training process using the set of evaluation data. The automatically tuned model parameters for the support vector machine may then be output directly from the training process. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method associated with a support vector machine having model parameters, comprising: receiving a set of evaluation data; automatically tuning, by a computer processor, model parameters during a training process using the set of evaluation data; and outputting, directly from the training process, the automatically tuned model parameters for the support vector machine. 2. The method of claim 1, wherein said automatic tuning is performed during a single phase of the training process. 3. The method of claim 1, further comprising: automatically rendering decisions using the support vector machine. 4. The method of claim 3, wherein the decisions are associated with at least one of: (i) classification, (ii) clustering, (iii) regression, (iv) anomaly detection, (v) association rules, (vi) reinforcement learning, (vii) structured prediction, (viii) feature learning, (ix) online learning, (x) semi-supervised learning, and (xi) grammar induction. 5. The method of claim 1, wherein said automatic tuning is performed by a set of i computational stations, where i is an integer greater than 1. 6. The method of claim 5, wherein said automatic tuning is performed by iteratively performing the following phases until convergence is achieved: a distribution phase to minimize a loss function associated with the support vector machine by distributing subsets of the evaluation data to the i computational stations; a collecting phase to enforce regularization and update shared model parameters; and a tuning phase that uses the set of evaluation data to update a trade-off parameter C. 7. The method of claim 6, wherein the distribution phase updates decoupled parameters, wherein w(i) represents a decision boundary for each subset of evaluation data x(i) distributed to an ith slave computational station as follows: w t + 1 ( ) = arg min w ( ) Cl ( α i w ( ) · φ ( x ( ) ) ) + 〈 α i , w ( ) - z 〉 + λ 2 w ( ) - z 2 8. The method of claim 6, wherein the collection phase collects distributed weighting parameters at a master computational station and updates regularization as follows: z t + 1 = arg min z β 1 2 z + ∑ i ( 〈 α t , w C ( ) - z 〉 + λ 2 w C ( ) - z 2 ) 9. The method of claim 6, wherein the tuning phase updates the trade-off parameter C as follows: C t + 1 = C t - γ ( ∑ ( x , y ) ∈ D v l ( yz · φ ( x ) ) + 2 β ∂ R ( C t ) ∂ ( C t ) ) 10. A non-transitory, computer-readable medium storing instructions that, when executed by a computer processor, cause the computer processor to perform a method associated with a support vector machine having model parameters, the method comprising: receiving a set of evaluation data; automatically tuning, by the computer processor, the model parameters during a training process using the set of evaluation data; and outputting, directly from the training process, the automatically tuned model parameters for the support vector machine. 11. The medium of claim 10, wherein said automatic tuning is performed during a single phase of the training process. 12. The medium of claim 10, wherein the method further comprises: automatically rendering decisions using the support vector machine, wherein the decisions are associated with at least one of: (i) classification, (ii) clustering, (iii) regression, (iv) anomaly detection, (v) association rules, (vi) reinforcement learning, (vii) structured prediction, (viii) feature learning, (ix) online learning, (x) semi-supervised learning, and (xi) grammar induction. 13. The medium of claim 10, wherein said automatic tuning is performed by a set of i computational stations, where i is an integer greater than 1, by iteratively performing the following phases until convergence is achieved: a distribution phase to minimize a loss function associated with the support vector machine by distributing subsets of the evaluation data to the i computational stations; a collecting phase to enforce regularization and update shared model parameters; and a tuning phase that uses the set of evaluation data to update a trade-off parameter C. 14. The medium of claim 13, wherein the distribution phase updates decoupled parameters, wherein w(i) represents a decision boundary for each subset of evaluation data x(i) distributed to an ith slave computational station as follows: w t + 1 ( ) = arg min w ( ) Cl ( α i w ( ) · φ ( x ( ) ) ) + 〈 α i , w ( ) - z 〉 + λ 2 w ( ) - z 2 15. The medium of claim 13, wherein the collection phase collects distributed weighting parameters at a master computational station and updates regularization as follows: z t + 1 = arg min z β 1 2 z + ∑ i ( 〈 α t , w C ( ) - z 〉 + λ 2 w C ( ) - z 2 ) 16. The medium of claim 13, wherein the tuning phase updates the trade-off parameter C as follows: C t + 1 = C t - γ ( ∑ ( x , y ) ∈ D v l ( yz · φ ( x ) ) + 2 β ∂ R ( C t ) ∂ ( C t ) ) 17. A system, comprising: a storage device to store a set of evaluation data; and a computer system coupled to the storage device to: (i) automatically tune the model parameters during a training process using the set of evaluation data, and (ii) output, directly from the training process, the automatically tuned model parameters for the support vector machine. 18. The system of claim 17, wherein said automatic tuning is performed during a single phase of the training process. 19. The system of claim 17, wherein the method further comprises: automatically rendering decisions using the support vector machine, wherein the decisions are associated with at least one of: (i) classification, (ii) clustering, (iii) regression, (iv) anomaly detection, (v) association rules, (vi) reinforcement learning, (vii) structured prediction, (viii) feature learning, (ix) online learning, (x) semi-supervised learning, and (xi) grammar induction. 20. The system of claim 17, wherein said automatic tuning is performed by a set of i computational stations, where i is an integer greater than 1, by iteratively performing the following phases until convergence is achieved: a distribution phase to minimize a loss function associated with the support vector machine by distributing subsets of the evaluation data to the i computational stations; a collecting phase to enforce regularization and update shared model parameters; and a tuning phase that uses the set of evaluation data to update a trade-off parameter C. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Some embodiments are associated with a support vector machine having model parameters. According to some embodiments, a set of evaluation data may be received and a computer processor may automatically tune the model parameters during a training process using the set of evaluation data. The automatically tuned model parameters for the support vector machine may then be output directly from the training process. |
|
G06N702 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Some embodiments are associated with a support vector machine having model parameters. According to some embodiments, a set of evaluation data may be received and a computer processor may automatically tune the model parameters during a training process using the set of evaluation data. The automatically tuned model parameters for the support vector machine may then be output directly from the training process. |
|
A capacity-analysis tool (CAT) provides a model framework for creating a model of a capacity-planning-target (CPT) system, e.g., a data center. The tool includes a model framework that, in turn, includes a closed-system template for creating CSMs, i.e., models of capacity-limited systems. A user uses the CAT to create CPT models using the CSMs as building blocks. A machine-learning engine is used to train the CPT model, converting parameter time-series data to functions of time. The trained CPT models are then used to make capacity-planning estimates, e.g., time remaining on a system before usage matches capacity. The CAT makes it easy to extend a model, e.g., by adding new dimensions (new factors of interest) in the form of new CSMs to which the new dimensions have been assigned. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A process comprising: for a capacity-planning target (CPT) system, creating a CPT model of the CPT system, the creating including creating plural closed-system models (CSMs), wherein each CSM is assigned a respective dimension and plural parameters for that dimension, at least one of the parameters being a capacity parameter, at least one of the parameters being a usage parameter or a demand parameter, at least one of the parameters corresponding to a respective time series of datapoints; training the CPT model using a machine-learning engine to convert the time-series to a respective function of time, the function of time thereby becoming part of the trained CPT model; and making, using the trained CPT model, capacity-analysis estimates based at least in part on the function of time. 2. The process of claim 1 wherein the making capacity-analysis estimates involves evaluating the function of time for a future time. 3. The process of claim 1 wherein the making capacity-analysis estimates involves combining plural functions of time across components and within a dimension to yield a system function of time for the dimension. 4. The process of claim 3 wherein the making capacity-analysis estimates further involves making estimates based at least in part on the system function of time. 5. The process of claim 4 wherein the making capacity-analysis estimates further involves combining estimates based at least in part on the system functions of time to yield the capacity-analysis estimates. 6. The process of claim 1 further comprising, after making the capacity analysis estimates, identifying a dimension not assigned to any CSM of the CPT model at the time the estimates were made; assigning the dimension to a new CSM and adding the new CSM to the CPT model; training the CPT model with the new CSM to yield a retrained CPT model; and making new capacity-analysis estimates using the retrained CPT model with the new CSM. 7. A capacity-analysis system comprising: a programmed hardware capacity-analysis tool including a model framework for creating, for a capacity-planning target (CPT) system, a CPT model for making capacity-planning estimates, the model framework including a closed-system template for creating closed-system models (CSMs) having associated therewith a dimension role to which a dimension can be assigned and parameter roles to which parameters can be assigned, the parameters including a capacity parameter and at least one of a usage parameter and a demand parameter, the CPT model including the CSMs; and a programmed hardware machine-learning engine for training the CPT model by converting time-series data associated with at least one of the parameters to at least one respective function of time. 8. The capacity-analysis system of claim 7 wherein the machine-learning engine is further to generate functions of time for some parameters based on functions of time generated from the time series. 9. The capacity-analysis system of claim 8 wherein the model framework includes algorithms, the CPT model including versions of these algorithms for making the capacity planning estimates. 10. The capacity-analysis system of claim 9 further comprising the CPT model. 11. The capacity analysis system of claim 10 wherein the algorithms provide for computing capacity-planning estimates for what-if scenarios provided to the CPT model. 12. A system comprising media encoded with code that, when executed by hardware, implements a process including: for a capacity-planning target (CPT) system, creating a CPT model of the CPT system, the CPT model including plural closed-system models (CSMs), wherein each CSM is assigned a respective dimension and plural parameters for that dimension, at least one of the parameters being a capacity parameter, at least one of the parameters being a usage parameter or a demand parameter, at least one of the parameters corresponding to a respective time series of datapoints; training the CPT model using a machine-learning engine to convert the at least one time-series to a respective function of time, the function of time thereby becoming part of the CPT model; and making capacity-analysis estimates, using the trained CPT model, based on the function of time. 13. The system of claim 12 wherein the making capacity-analysis estimates includes evaluating the at least one function of time for a future time. 14. The system of claim 12 wherein the making capacity-analysis estimates includes combining plural functions of time across components and within a dimension to yield a system function of time for the dimension. 15. The system of claim 14 wherein the making capacity-analysis estimates further involves making estimates based at least in part on the system function of time. 16. The system of claim 15 wherein the making capacity-analysis estimates further includes combining the estimates based on system functions of time to yield the capacity-analysis estimates. 17. The system of claim 12 further comprising, after making the capacity analysis estimates, identifying a dimension not assigned to any CSM of the CPT model at the time the estimates were made; assigning the dimension to a new CSM and adding the new CSM to the CPT model; training the CPT model with the new CSM to yield a retrained CPT model; and making new capacity-analysis estimates using the retrained CPT model with the new CSM. 18. The system of claim 12 further comprising the hardware. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A capacity-analysis tool (CAT) provides a model framework for creating a model of a capacity-planning-target (CPT) system, e.g., a data center. The tool includes a model framework that, in turn, includes a closed-system template for creating CSMs, i.e., models of capacity-limited systems. A user uses the CAT to create CPT models using the CSMs as building blocks. A machine-learning engine is used to train the CPT model, converting parameter time-series data to functions of time. The trained CPT models are then used to make capacity-planning estimates, e.g., time remaining on a system before usage matches capacity. The CAT makes it easy to extend a model, e.g., by adding new dimensions (new factors of interest) in the form of new CSMs to which the new dimensions have been assigned. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A capacity-analysis tool (CAT) provides a model framework for creating a model of a capacity-planning-target (CPT) system, e.g., a data center. The tool includes a model framework that, in turn, includes a closed-system template for creating CSMs, i.e., models of capacity-limited systems. A user uses the CAT to create CPT models using the CSMs as building blocks. A machine-learning engine is used to train the CPT model, converting parameter time-series data to functions of time. The trained CPT models are then used to make capacity-planning estimates, e.g., time remaining on a system before usage matches capacity. The CAT makes it easy to extend a model, e.g., by adding new dimensions (new factors of interest) in the form of new CSMs to which the new dimensions have been assigned. |
|
A quantum computer comprises of at least one qubit formed from holes created with acceptor atoms (10) in crystalline silicon (12) and a pair of gates (14, 16) located above the acceptor atoms (10) to apply direct electric field and alternating electric field for switching, manipulating the qubit such that quantum information resulting from being manipulated is stored from decoherence. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A quantum computer, comprising: one or more electrically active acceptor atoms incorporated into a crystalline silicon, wherein each acceptor atom creates a single hole bound to the acceptor atom within the crystalline silicon to form an acceptor qubit, wherein each acceptor qubit is either in an ‘ON’ state where the acceptor qubit is able to be manipulated by applying an alternating electric field, or in an ‘OFF’ state where the qubit is protected from decoherence; a first pair of gates located above each acceptor atom, by which the alternating electric field applied to the crystalline silicon manipulates a spin of the hole of the acceptor atom at resonance, and by which a direct electric field is applied to the crystalline silicon to control an interaction strength between one or more qubit levels, to selectively switch the acceptor qubit between an ‘ON’ state in which quantum information can be manipulated and an ‘OFF’ state in which quantum information, resulting from being manipulated, is stored in basis that is protected from decoherence, and wherein the first pair of gates are also used to readout results of quantum computations as charge signals that represent one or more spin states. 2. The quantum computer as claimed in claim 1, wherein the spin states are sensed by a nearby single electron transistor or a quantum point contacted. 3. The quantum computer as claimed in claim 1, wherein a source of external magnetic field is applied to the acceptor atom to lift a four-fold degeneracy of a ground state of the acceptor atom to facilitate flips of the spin of the hole at resonance using the alternating electric field. 4. The quantum computer as claimed in claim 1, wherein two lowest Zeeman levels may be used as working levels when the acceptor qubit is in the ‘ON’ state. 5. The quantum computer as claimed in claim 4, wherein coherent qubit rotations are driven by applying the alternating electric field to the gates in resonance with the two lowest Zeeman levels. 6. The quantum computer as claimed in claim 1, wherein the direct electric field is applied perpendicular to an interface to switch the qubit to the ‘OFF’ state by bringing the hole bound to the acceptor atom closer to the interface. 7. The quantum computer as claimed in claim 1, wherein a transfer of quantum information from the ‘ON’ state to the ‘OFF’ state involves mixing working levels. 8. The quantum computer as claimed in claim 7, wherein the transfer of quantum information involves simultaneously changing both a magnetic field parallel to the interface and the direct electric field. 9. The quantum computer as claimed in claim 8, wherein the transfer of quantum information is required to be completed faster than the spin decoherence time. 10. A method for operating a quantum computer comprising one or more electrically active acceptor atoms within a crystalline silicon, comprising the steps of: applying an alternating electric field to the crystalline silicon to coherently drive rotations of a qubit at resonance; applying a direct electric field to the crystalline silicon to control the interaction strength and to selectively switch the qubit between an ‘ON’ state in which quantum information can be manipulated, and an ‘OFF’ state in which quantum information, resulting from being manipulated, is stored in basis that is protected from decoherence; and reading out results of quantum computations as charge signals that represent spin states. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: A quantum computer comprises of at least one qubit formed from holes created with acceptor atoms (10) in crystalline silicon (12) and a pair of gates (14, 16) located above the acceptor atoms (10) to apply direct electric field and alternating electric field for switching, manipulating the qubit such that quantum information resulting from being manipulated is stored from decoherence. |
|
G06N99002 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A quantum computer comprises of at least one qubit formed from holes created with acceptor atoms (10) in crystalline silicon (12) and a pair of gates (14, 16) located above the acceptor atoms (10) to apply direct electric field and alternating electric field for switching, manipulating the qubit such that quantum information resulting from being manipulated is stored from decoherence. |
|
Methods for providing data analysis service by a service provider to a data owner are described. The data owner transmits training data to the data analysis service provider, and the latter computes a model from the training data. In one method, the service provider transmits the model back to the data owner, which uses the model to generate predictions from prediction input. In another method, the data owner further transmits prediction input to the service provider, and the latter uses the computed model and the prediction input to generate predictions and then transmits the predictions back to the data owner. Prior to transmitting the training data and the prediction input, the data owner performs variable name anonymization and a variable transformation on the training data and prediction data point to obscure the meaning of the variables in the data. This prevents possible misuse of the data owner's data by unauthorized parties. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method implemented in a first server operated by a data owner and a second server operated by a data analysis service provider, comprising: (a) the first server transmitting training data to the second server; (b) the second server analyzing the training data received from the first server using machine learning to develop a model; (c) the first server transmitting a prediction input to the second server; (d) the second server computing a prediction using the model developed in step (b) and the prediction input received from the first server; and (e) the second server transmitting the prediction to the first server. 2. The method of claim 1, further comprising, before step (a): (f) the first server obtaining data to be analyzed, the data including a plurality of data points, each data point including a plurality of variables each having a value; and (g) the first server pre-processing the data, including performing a variable transformation on each data point, to generate pre-processed data, wherein the pre-processed data and the data to be analyzed have different variable value distributions; wherein in step (a), the first server transmits the pre-processed data as the training data to the second server; the method further comprising, before step (c): (h) the first server pre-processing a prediction data point, the prediction data point including the plurality of variables each having a value, the pre-processing including performing the variable transformation on the prediction data point to generate pre-processed prediction data point; wherein in (c), the first server transmits the pre-processed prediction data point as the prediction input to the second server. 3. The method of claim 1, further comprising, before step (a): (f) the first server obtaining data to be analyzed, the data including a plurality of data points, each data point including a first plurality of variables each having a value; (g) the first server pre-processing the data, including performing a variable transformation on each data point, to generate pre-processed data in which each data point includes a second plurality of variables each having a value, wherein at least one variable x, among the first plurality of variables is not among the second plurality of variables, and a set of replacement variables Zs to Zt among the second plurality of variables are not among the first plurality of variables; wherein in step (a), the first server transmits the pre-processed data as the training data to the second server; the method further comprising, before step (c): (h) the first server pre-processing a prediction data point, the prediction data point including the first plurality of variables each having a value, the pre-processing including performing the variable transformation on the prediction data point to generate pre-processed prediction data point which includes the second plurality of variables each having a value; wherein in (c), the first server transmits the pre-processed prediction data point as the prediction input to the second server. 4. The method of claim 3, wherein the variable transformation in the pre-processing steps (g) and (h) includes: for the first variable Xj, defining the set of replacement variables Zs to Zt which satisfy the condition: Xj=λ0+λsZs+ . . . +λtZt wherein λ0, λs, . . . , λt are a set of coefficients, and wherein values of the set of replacement variables are dependent on the value of the first variable and/or auxiliary information, the auxiliary information being known to the first server but unknown to the second server. 5. A method implemented in a first server operated by a data owner and a second server operated by a data analysis service provider, comprising: (a) the first server transmitting training data to the second server; (b) the second server analyzing the training data received from the first server using machine learning to develop a model; (c) the second server transmitting the model to the first server; and (d) the first server computing a prediction using the model received from the second server and a prediction input. 6. The method of claim 5, further comprising, before step (a): (e) the first server obtaining data to be analyzed, the data including a plurality of data points, each data point including a plurality of variables each having a value; and (f) the first server pre-processing the data, including performing a variable transformation on each data point, to generate pre-processed data, wherein the pre-processed data and the data to be analyzed have different variable value distributions; wherein in step (a), the first server transmits the pre-processed data as the training data to the second server; the method further comprising, before step (d): (g) the first server pre-processing a prediction data point, the prediction data point including the plurality of variables each having a value, the pre-processing including performing the variable transformation on the prediction data point to generate pre-processed prediction data point; wherein in (d), the first server uses the pre-processed prediction data point as the prediction input. 7. The method of claim 5, further comprising, before step (a): (e) the first server obtaining data to be analyzed, the data including a plurality of data points, each data point including a first plurality of variables each having a value; (f) the first server pre-processing the data, including performing a variable transformation on each data point, to generate pre-processed data in which each data point includes a second plurality of variables each having a value, wherein at least one variable Xj among the first plurality of variables is not among the second plurality of variables, and a set of replacement variables Zs to Zt among the second plurality of variables are not among the first plurality of variables; wherein in step (a), the first server transmits the pre-processed data as the training data to the second server; the method further comprising, before step (d): (g) the first server pre-processing a prediction data point, the prediction data point including the first plurality of variables each having a value, the pre-processing including performing the variable transformation on the prediction data point to generate pre-processed prediction data point which includes the second plurality of variables each having a value; wherein in (d), the first server transmits the pre-processed prediction data point as the prediction input to the second server. 8. The method of claim 7, wherein the variable transformation in the pre-processing steps (f) and (g) includes: for the first variable Xj, defining the set of replacement variables Zs to Zt which satisfy the condition: Xj=λ0+λsZs+ . . . λtZt wherein λ0, λs, . . . , λt are a set of coefficients, and wherein values of the set of replacement variables are dependent on the value of the first variable and/or auxiliary information, the auxiliary information being known to the first server but unknown to the second server. 9. A method implemented in a first server operated by a data owner, the first server cooperating with a second server operated by a data analysis service provider, the method comprising: (a) obtaining data to be analyzed, the data including a plurality of data points, each data point including a first plurality of variables each having a value; (b) pre-processing the data, including performing a variable transformation on each data point, to generate pre-processed data in which each data point includes a second plurality of variables each having a value, wherein at least one variable Xj among the first plurality of variables is not among the second plurality of variables, and a set of replacement variables Zs to Zt among the second plurality of variables are not among the first plurality of variables; (c) transmitting the training data to the second server; and (d) pre-processing a prediction data point, the prediction data point including the first plurality of variables each having a value, the pre-processing including performing the variable transformation on the prediction data point to generate pre-processed prediction data point which includes the second plurality of variables each having a value. 10. The method of claim 9, further comprising: (e) transmitting the pre-processed prediction data point as prediction input to the second server; and (f) receiving a prediction from the second server which has been computed by the second server based on the training data and the prediction input. 11. The method of claim 9, further comprising: (e) receiving a model from the second server which has been learned by the second server from the training data; and (f) computing a prediction using the model received from the second server and the pre-processed prediction data point as prediction input. 12. The method of claim 9, wherein the variable transformation in the pre-processing steps (b) and (d) includes: for the first variable Xj, defining the set of replacement variables Zs to Zt which satisfy the condition: Xj=λ0+λsZs+ . . . λtZt wherein λ0, λs, . . . λt are a set of coefficients, and wherein values of the set of replacement variables are dependent on the value of the first variable and/or auxiliary information, the auxiliary information being known to the first server but unknown to the second server. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods for providing data analysis service by a service provider to a data owner are described. The data owner transmits training data to the data analysis service provider, and the latter computes a model from the training data. In one method, the service provider transmits the model back to the data owner, which uses the model to generate predictions from prediction input. In another method, the data owner further transmits prediction input to the service provider, and the latter uses the computed model and the prediction input to generate predictions and then transmits the predictions back to the data owner. Prior to transmitting the training data and the prediction input, the data owner performs variable name anonymization and a variable transformation on the training data and prediction data point to obscure the meaning of the variables in the data. This prevents possible misuse of the data owner's data by unauthorized parties. |
|
G06N7005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Methods for providing data analysis service by a service provider to a data owner are described. The data owner transmits training data to the data analysis service provider, and the latter computes a model from the training data. In one method, the service provider transmits the model back to the data owner, which uses the model to generate predictions from prediction input. In another method, the data owner further transmits prediction input to the service provider, and the latter uses the computed model and the prediction input to generate predictions and then transmits the predictions back to the data owner. Prior to transmitting the training data and the prediction input, the data owner performs variable name anonymization and a variable transformation on the training data and prediction data point to obscure the meaning of the variables in the data. This prevents possible misuse of the data owner's data by unauthorized parties. |
|
A multi-pattern matching algorithm may be provided that includes: a moving step of moving a moving window from the start of a string one byte by one byte; a DF1 checking step of converting the string on a current position of the moving window into an integer value, and of checking whether or not a bit of a related position in a first direct filter DF1 for patterns having lengths larger than 2 bytes is set to 1; a DF moving step of checking one or more direct filters DF when the bit is set to 1 according to the DF1 checking step; a re-moving step of moving the moving window by one byte again when the bit of a related position in the direct filter DF, which has been checked lastly, is 0; and a terminating step of checking whether the moving window is located at the end of the string or not, and of terminating the algorithm when the moving window is positioned at the end of the string. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A multi-pattern matching algorithm comprising: a moving step of moving a moving window from the start of a string one byte by one byte; a DF1 checking step of converting the string on a current position of the moving window into an integer value, and of checking whether or not a bit of a related position in a first direct filter DF1 for patterns having lengths larger than 2 bytes is set to 1; a DF moving step of checking one or more direct filters DF when the bit is set to 1 according to the DF1 checking step; a re-moving step of moving the moving window by one byte again when the bit of a related position in the direct filter DF, which has been checked lastly, is 0; and a terminating step of checking whether the moving window is located at the end of the string or not, and of terminating the algorithm when the moving window is positioned at the end of the string. 2. The multi-pattern matching algorithm of claim 1, further comprising, after the DF moving step, a DF4 checking step of checking whether or not a bit of a related position in a fourth direct filter DF4 for patterns having lengths larger than 2 bytes and less than 4 bytes is set to 1. 3. The multi-pattern matching algorithm of claim 2, further comprising, after the DF4 checking step, a PID recording step of when the bit of a related position in the fourth direct filter DF4 is set to 1, recording a pattern ID (PID) corresponding to the string in which the moving window is located, with reference to a first compact table CT1 storing PIDs of the patterns having lengths larger than 2 bytes and less than 4 bytes. 4. The multi-pattern matching algorithm of claim 1, further comprising, after the DF moving step, a DF2 checking step of moving the moving window by two bytes from the current position, of converting the string of a length as much as 2 bytes on the moved position into an integer value, and of checking whether or not a bit of a related position in a second direct filter DF2 for patterns having lengths larger than 4 bytes is set to 1. 5. The multi-pattern matching algorithm of claim 4, further comprising, after the DF2 checking step, a DF5 checking step of, when the bit of a related position in the second direct filter DF2 is 1, checking whether or not a bit of a related position in a fifth direct filter DF5 for patterns having lengths larger than 4 bytes and less than 8 bytes is set to 1. 6. The multi-pattern matching algorithm of claim 5, further comprising, after the DF5 checking step, a PID recording step of when the bit of a related position in the fifth direct filter DF5 is 1, checking whether or not a pattern ID (PID) corresponding to the string in which the moving window is located, with reference to a second compact table CT2 storing PIDs of the patterns having lengths larger than 4 bytes and less than 8 bytes, and of when the PID corresponding to the string exists, recording the PID. 7. The multi-pattern matching algorithm of claim 1, further comprising, after the DF moving step, a DF3 checking step of moving the moving window by six bytes from the current position, of converting the string of a length as much as 2 bytes on the moved position into an integer value, and of checking whether or not a bit of a related position in a third direct filter DF3 for patterns having lengths larger than 8 bytes is set to 1. 8. The multi-pattern matching algorithm of claim 7, further comprising, after the DF3 checking step, a PID recording step of, the bit of a related position in the third direct filter DF3 is set to 1, recording a pattern ID (PID) corresponding to the string in which the moving window is located, with reference to a third compact table CT3 storing PIDs of the patterns having lengths larger than 8 bytes. 9. The multi-pattern matching algorithm of claim 1, wherein the algorithm is used in a network intrusion detection system (NIDS). 10. A program which is stored in a medium and performs: a moving step of moving a moving window from the start of a string one byte by one byte; a DF1 checking step of converting the string on a current position of the moving window into an integer value, and of checking whether or not a bit of a related position in a first direct filter DF1 for patterns having lengths larger than 2 bytes is set to 1; a DF moving step of moving the moving window to one or more direct filters DF when the bit is set to 1 according to the DF1 checking step; a re-moving step of moving the moving window by one byte again when the bit of a related position in the direct filter DF, which has been checked lastly, is 0; and a terminating step of checking whether the moving window is located at the end of the string or not, and of terminating the algorithm when the moving window is positioned at the end of the string. 11. A multi-pattern matching processing device comprising: a direct filter DF which is a bit array having a plurality of bits, each of which indicates whether one or more consecutive ASCII codes corresponding to its index belongs to a portion of a particular pattern or not, and is composed of one or more direct filters, each of which has information on 2n (n=0, 1, 2, 3, . . . )-th two bytes of the pattern according to a length of the pattern; and at least one compact table CT which is a structure for recording pattern IDs of the patterns existing in a string and for finding out what pattern exists in the string, and stores the pattern ID according to pattern groups formed based on the length of the pattern. 12. The multi-pattern matching processing device of claim 11, wherein the direct filter DF comprises a first direct filter DF1 comprising information on the two headmost bytes of all of the patterns, a second direct filter DF2 comprising information on the second two bytes of the patterns having lengths larger than 4 bytes, a third direct filter DF3 comprising information on the fourth two bytes of the patterns having lengths larger than 8 bytes, and a fourth direct filter DF4 comprising information on the two headmost bytes of the patterns having lengths larger than 2 bytes and less than 4 bytes. 13. The multi-pattern matching processing device of claim 12, wherein the direct filter DF further comprises a fifth direct filter DF5 comprising information on the second two bytes of the patterns having lengths larger than 4 bytes and less than 8 bytes. 14. The multi-pattern matching processing device of claim 11, wherein the compact table CT comprises a first compact table CT1 comprising the pattern IDs of the patterns having lengths larger than 2 bytes and less than 4 bytes, a second compact table CT2 comprising the pattern IDs of the patterns having lengths larger than 4 bytes and less than 8 bytes, and a third compact table CT3 comprising the pattern IDs of the patterns having lengths larger than 8 bytes. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A multi-pattern matching algorithm may be provided that includes: a moving step of moving a moving window from the start of a string one byte by one byte; a DF1 checking step of converting the string on a current position of the moving window into an integer value, and of checking whether or not a bit of a related position in a first direct filter DF1 for patterns having lengths larger than 2 bytes is set to 1; a DF moving step of checking one or more direct filters DF when the bit is set to 1 according to the DF1 checking step; a re-moving step of moving the moving window by one byte again when the bit of a related position in the direct filter DF, which has been checked lastly, is 0; and a terminating step of checking whether the moving window is located at the end of the string or not, and of terminating the algorithm when the moving window is positioned at the end of the string. |
|
G06N5047 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A multi-pattern matching algorithm may be provided that includes: a moving step of moving a moving window from the start of a string one byte by one byte; a DF1 checking step of converting the string on a current position of the moving window into an integer value, and of checking whether or not a bit of a related position in a first direct filter DF1 for patterns having lengths larger than 2 bytes is set to 1; a DF moving step of checking one or more direct filters DF when the bit is set to 1 according to the DF1 checking step; a re-moving step of moving the moving window by one byte again when the bit of a related position in the direct filter DF, which has been checked lastly, is 0; and a terminating step of checking whether the moving window is located at the end of the string or not, and of terminating the algorithm when the moving window is positioned at the end of the string. |
|
An individual neuron circuit calculates a first value based on a sum of products each obtained by multiplying one of weight values, each representing connection or disconnection between a corresponding neuron circuit and one of the other neuron circuits, by a corresponding one of output signals of the other neuron circuits and outputs 0 or 1, based on a result of comparison between a second value obtained by adding a noise value to the first value and a threshold. An arbitration circuit allows, when first output signals of first neuron circuits interconnected among the neuron circuits simultaneously change based on the weight values, updating of only one of the first output signals of the first neuron circuits and allows, when second output signals of second neuron circuits not interconnected simultaneously change, updating of the second output signals. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. An ising device comprising: a plurality of neuron circuits which each calculate a first value based on a sum of products each obtained by multiplying one of a plurality of weight values each representing connection or disconnection between a corresponding neuron circuit and one of others of the plurality of neuron circuits, by a corresponding one of a plurality of output signals of the others of the plurality of neuron circuits, and output 0 or 1, based on a result of comparison between a second value obtained by adding a noise value to the first value and a threshold; and an arbitration circuit which allows, when a plurality of first output signals of a plurality of first neuron circuits interconnected among the plurality of neuron circuits simultaneously change based on the plurality of weight values, updating of only one of the first output signals of the plurality of first neuron circuits and allows, when a plurality of second output signals of a plurality of second neuron circuits not interconnected simultaneously change, updating of the plurality of second output signals. 2. The ising device according to claim 1, wherein the arbitration circuit generates first identification information that identifies one of the plurality of first neuron circuits, and wherein, among the plurality of neuron circuits, a third neuron circuit corresponding to the first identification information performs updating, based on the plurality of output signals. 3. The ising device according to claim 2, comprising: a memory that holds a group of items of identification information that respectively identifies the plurality of neuron circuits, wherein the arbitration circuit reads the group of items of identification information from the memory, extracts a group of items of first identification information that respectively identifies the plurality of first neuron circuits from the group of items of identification information, and randomly selects the first identification information from the group of items of first identification information or selects first identification information having a maximum or minimum value from the group of items of first identification information. 4. The ising device according to claim 3, wherein the arbitration circuit extracts second identification information that identifies a fourth neuron circuit connected to the third neuron circuit based on the plurality of weight values from the group of items of identification information and excludes the second identification information from the group of items of first identification information. 5. A control method of an ising device, the method comprising: for the ising device including a plurality of neuron circuits which each calculate a first value based on a sum of products each obtained by multiplying one of a plurality of weight values each representing connection or disconnection between a corresponding neuron circuit and one of others of the plurality of neuron circuits, by a corresponding one of a plurality of output signals of the others of the plurality of neuron circuits and output 0 or 1, based on a result of comparison between a second value obtained by adding a noise value to the first value and a threshold, and an arbitration circuit which allows, when a plurality of first output signals of a plurality of first neuron circuits interconnected among the plurality of neuron circuits simultaneously change based on the plurality of weight values, updating of only one of the first output signals of the plurality of first neuron circuits and allows, when a plurality of second output signals of a plurality of second neuron circuits not interconnected simultaneously change, updating of the plurality of second output signals, setting, by a control device coupled to the ising device, the plurality of weight values; causing, by the control device, a random signal generation circuit to randomly select the plurality of first neuron circuits or the plurality of second neuron circuits to be enabled among the plurality of neuron circuits; and controlling, by the control device, an amplitude of the noise value outputted by a noise generation circuit. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: An individual neuron circuit calculates a first value based on a sum of products each obtained by multiplying one of weight values, each representing connection or disconnection between a corresponding neuron circuit and one of the other neuron circuits, by a corresponding one of output signals of the other neuron circuits and outputs 0 or 1, based on a result of comparison between a second value obtained by adding a noise value to the first value and a threshold. An arbitration circuit allows, when first output signals of first neuron circuits interconnected among the neuron circuits simultaneously change based on the weight values, updating of only one of the first output signals of the first neuron circuits and allows, when second output signals of second neuron circuits not interconnected simultaneously change, updating of the second output signals. |
|
G06N3049 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: An individual neuron circuit calculates a first value based on a sum of products each obtained by multiplying one of weight values, each representing connection or disconnection between a corresponding neuron circuit and one of the other neuron circuits, by a corresponding one of output signals of the other neuron circuits and outputs 0 or 1, based on a result of comparison between a second value obtained by adding a noise value to the first value and a threshold. An arbitration circuit allows, when first output signals of first neuron circuits interconnected among the neuron circuits simultaneously change based on the weight values, updating of only one of the first output signals of the first neuron circuits and allows, when second output signals of second neuron circuits not interconnected simultaneously change, updating of the second output signals. |
|
A computer-implementable method for managing a cognitive graph comprising: receiving data from a plurality of data sources; processing the data from the plurality of data sources, the processing the data from the plurality of data sources identifying a plurality of knowledge elements; and, storing the knowledge elements within the cognitive graph as a collection of knowledge elements, the storing universally representing knowledge obtained from the data. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implementable method for managing a cognitive graph comprising: receiving data from a plurality of data sources; processing the data from the plurality of data sources, the processing the data from the plurality of data sources identifying a plurality of knowledge elements; and, storing the knowledge elements within the cognitive graph as a collection of knowledge elements, the storing universally representing knowledge obtained from the data. 2. The method of claim 1, further comprising: generating a cognitive insight based upon the collection of knowledge elements stored within the cognitive graph. 3. The method of claim 1, further comprising: identifying at least some knowledge elements within the collection of knowledge elements as at least one of facts, opinions, descriptions, and skills. 4. The method of claim 1, further comprising: identifying at least some knowledge elements within the collection of knowledge elements as at least one of statements, assertions, beliefs, perceptions, preferences, sentiments, attitudes and opinions and associating at least some knowledge elements with an entity responsible for generating the at least one of statements, assertions, beliefs, perceptions, preferences, sentiments, attitudes and opinions. 5. The method of claim 1, wherein: the cognitive graph comprises an entailment graph, the entailment graph modeling knowledge through inheritance. 6. The method of claim 1, wherein: the knowledge elements are stored within the cognitive graph as nodes; and, subsets of nodes are related via edges. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A computer-implementable method for managing a cognitive graph comprising: receiving data from a plurality of data sources; processing the data from the plurality of data sources, the processing the data from the plurality of data sources identifying a plurality of knowledge elements; and, storing the knowledge elements within the cognitive graph as a collection of knowledge elements, the storing universally representing knowledge obtained from the data. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A computer-implementable method for managing a cognitive graph comprising: receiving data from a plurality of data sources; processing the data from the plurality of data sources, the processing the data from the plurality of data sources identifying a plurality of knowledge elements; and, storing the knowledge elements within the cognitive graph as a collection of knowledge elements, the storing universally representing knowledge obtained from the data. |
|
A computer-implemented method for profiling a population of examples includes a computer system creating a rule collection comprising a plurality of rules, wherein each rule describes a respective corresponding sub-population of the examples according to a conjunction of a plurality of feature-value pairs. The computer system generates a precisely descriptive profile by performing a search process on the rule collection to identify a rule that either maximizes or minimizes the value of a user-specified target feature in the respective corresponding sub-population. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method for profiling a population of examples, the method comprising: creating, by a computer system, a rule collection comprising a plurality of rules, wherein each rule describes a respective corresponding sub-population of the examples according to a conjunction of a plurality of feature-value pairs; and generating, by the computer system, a precisely descriptive profile by performing a search process on the rule collection to identify a rule that either maximizes or minimizes a value of a user-specified target feature in the respective corresponding sub-population. 2. The method of claim 1, wherein the search process is implemented using a beam search algorithm. 3. The method of claim 1, wherein the search process is implemented using a Monte Carlo search algorithm. 4. The method of claim 1, wherein the search process maximizes a utility measurement for each rule in the plurality of rules. 5. The method of claim 4, wherein the utility measurement is based on a deviation (above or below) of the user-specified target feature in the respective corresponding sub-population from the mean value of the user-specified target feature in the population of examples. 6. The method of claim 5, wherein the utility measurement is further based on a weighted function of the value corresponding to the user-specified target feature and a sub-population count proscribed by the rule. 7. The method of claim 4, wherein the utility measurement is the magnitude of the Z-score of the respective corresponding sub-population, implicitly defining a weighting between population count and a target feature deviation from the mean. 8. The method of claim 4, wherein the utility measurement includes a constraint selected from (i) a first constraint that the respective sub-population must include a minimum number of population members or (ii) a second constraint that the respective corresponding sub-population must comprise a minimum percentage of the population. 9. The method of claim 1 wherein the number of feature-value pairs in the plurality of rules is bounded by a user-specified parameter. 10. The method of claim 1, further comprising: prior to creating the rule collection, performing a pre-processing process on the population examples comprising: identifying a plurality of ordinal features included in the population of examples which correspond to the user-specified target feature; dividing the plurality of ordinal features into a plurality of bins according to corresponding feature values; and performing a condition creation process for each rule comprising: identifying a subset of the plurality of bins having a significant deviation from the mean value of the population with respect to the user-specified target feature, and combining ordinal features included in the subset of the plurality of bins. 11. The method of claim 10, wherein the pre-processing process further comprises: identifying a plurality of nominal features included in the population of examples; and during the condition creation process for each rule, combining the plurality of nominal features into disjunctive subsets of the population of examples. 12. The method of claim 1, wherein the method further comprises an iterative process comprising: removing a particular sub-population covered from by the precisely descriptive profile from an example collection; and repeating the search process on remaining examples in the example collection to generate a second precisely descriptive profile. 13. A system for profiling a population of examples, the system comprising: a database configured to store a rule collection comprising a plurality of rules, wherein each rule describes a respective corresponding sub-population of the examples according to a conjunction of a plurality of feature-value pairs; and a plurality of processors configured to generate a precisely descriptive profile by performing a search process on the rule collection to identify a rule that either maximizes or minimizes a value of a user-specified target feature in the respective corresponding sub-population. 14. A computer-implemented method for profiling a population of examples, the method comprising: receiving, by a computer system, a user-specified target feature; determining, by the computer system, a performance measurement for each example in the population with regards to the user-specified target feature; identifying, by the computer system, a sub-population of the examples based on the performance measurement determined for each example, wherein the sub-population comprises one of (i) highest performers with respect to the user-specified target feature or (ii) lowest performers with respect to the user-specified target feature; determining, by the computer system, a population mean value for the user-specified target feature across the population; identifying, by the computer system, feature-value pairs from the sub-population that deviate from the population mean value by more than a predetermined threshold value; and displaying the identified feature-value pairs. 15. The method of claim 14, further comprising: performing similarity-based clustering on the sub-population to generate a plurality of mutually exclusive sets; for each mutually exclusive set, determining a first deviation value indicative of a degree to which the mutually exclusive set deviates from the population mean value with respect to the user-specified target feature; and displaying the first deviation value associated with each of the plurality of mutually exclusive sets. 16. The method of claim 15, further comprising: for each mutually exclusive set in the plurality of mutually exclusive sets, determining a second deviation value indicative of a degree to which the mutually exclusive set deviates from other members of the plurality of mutually exclusive sets with respect to the user-specified target feature; and displaying the second deviation value associated with each of the plurality of mutually exclusive sets. 17. The method of claim 15, wherein the plurality of mutually exclusive sets are produced hierarchically on the sub-population. 18. The method of claim 15, wherein the similarity-based clustering produces a quasi-optimal number of mutually exclusive sets by an iterative process comprising: creating a new set; and successively adding clusters to the new set until the new set does not significantly differ from one or more prior sets. 19. The method of claim 15, wherein the computer system comprises a plurality of processors and the similarity-based clustering is performed in parallel. 20. The method of claim 14, wherein the computer system comprises a plurality of processors and each processor is configured to operate on a subset of the population in order to identify examples in the subset of the population meeting predetermined performance criteria. 21. The method of claim 14, wherein the computer system comprises a plurality of processors and each processor is configured to determine cohort deviation values over successive slices of the population in parallel. 22. The method of claim 14, further comprising: identifying, by the computer system, a plurality of cohorts in the population related to the user-specified target feature; identifying, by the computer system, additional feature-value pairs from the plurality of cohorts that deviate from the population mean value by more than the predetermined threshold value; and displaying the additional feature-value pairs. 23. A system for profiling a population of examples, the system comprising: a network interface configured to receive a user-specified target feature; a plurality of processors configured to: determine a performance measurement for each example in the population with regards to the user-specified target feature, identify a sub-population of the examples based on the performance measurement determined for each example, wherein the sub-population comprises one of (i) highest performers with respect to the user-specified target feature or (ii) lowest performers with respect to the user-specified target feature, determine a population mean value for the user-specified target feature across the population, and identify feature-value pairs from the sub-population that deviate from the population mean value by more than a predetermined threshold value; and a display configured to present the identified feature-value pairs. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A computer-implemented method for profiling a population of examples includes a computer system creating a rule collection comprising a plurality of rules, wherein each rule describes a respective corresponding sub-population of the examples according to a conjunction of a plurality of feature-value pairs. The computer system generates a precisely descriptive profile by performing a search process on the rule collection to identify a rule that either maximizes or minimizes the value of a user-specified target feature in the respective corresponding sub-population. |
|
G06N5025 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A computer-implemented method for profiling a population of examples includes a computer system creating a rule collection comprising a plurality of rules, wherein each rule describes a respective corresponding sub-population of the examples according to a conjunction of a plurality of feature-value pairs. The computer system generates a precisely descriptive profile by performing a search process on the rule collection to identify a rule that either maximizes or minimizes the value of a user-specified target feature in the respective corresponding sub-population. |
|
Embodiments associate a relevant semantic data type (e.g., date) with incoming raw data (e.g., a column of digits) which lacks metadata. Assignment of semantic data type is inferred from a plurality of features. A first step determines a first feature comprising success rate in converting the raw data into various semantic data types. Then, alignment between observed/reference distributions of other features (e.g., data first digit, data length) is determined per-semantic data type. Total scores for each semantic data type are calculated from the combined features, and used as a basis for ranking the semantic data types. The total scores may reflect a weighting of the various features. In a second step, top-ranked semantic data types may be further differentiated from one another by applying additional features. User feedback regarding accuracy of semantic data type assignment, may be incorporated into training data used to modify the feature reference distributions. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-implemented method comprising: receiving a plurality of data points lacking associated metadata; determining a first score of a first feature comprising a success rate converting the plurality of data points to a semantic data type; determining a second score from alignment between an observed distribution of a second feature in the plurality of data points, and a reference distribution of the second feature in the semantic data type; determining a third score from alignment between an observed distribution of a third feature in the plurality of data points, and a reference distribution of the third feature in the semantic data type; calculating a total score for the semantic data type from the first, second, and third scores; determining an initial ranking of the total score in comparison with another total score reflecting the first, second, and third features in another semantic data type; identifying a fourth feature differentiating between the semantic data type and the other semantic data type; assigning a final ranking to the total score and the other total score by, determining alignment between an observed distribution of the fourth feature in the plurality of data points, and reference distributions of the fourth feature in the semantic data type and in the other semantic data type; assigning the semantic data type to the plurality of data points based upon the final ranking; and storing the plurality of data points with metadata indicating the semantic data type. 2. A method as in claim 1 wherein: the alignment between the observed distribution of the second feature and the reference distribution of the second feature is reflected by a Kolmogorov-Smirnov distance. 3. A method as in claim 1 wherein the second feature comprises a data point length. 4. A method as in claim 1 wherein the second feature comprises a data point first digit. 5. A method as in claim 1 wherein the fourth feature comprises a number of digits behind a decimal point. 6. A method as in claim 1 wherein the fourth feature comprises a data point digit other than a first digit. 7. A method as in claim 1 further comprising: deriving the reference distribution of the second feature from training data; receiving feedback regarding accuracy of the assignment of semantic data type; and updating the training data to reflect the feedback. 8. A method as in claim 1 further comprising: receiving feedback regarding accuracy of the assignment of semantic data type; calculating the total score by assigning respective weight coefficients to the first, second, and third scores; and updating the respective weight coefficients based upon the feedback. 9. A method as in claim 1 wherein converting the plurality of data points to the semantic data type employs a pattern matching mechanism comprising regular expressions specifying class-specific search patterns. 10. A non-transitory computer readable storage medium embodying a computer program for performing a method, said method comprising: receiving a plurality of data points lacking associated metadata; determining a first score of a first feature comprising a success rate converting the plurality of data points to a semantic data type; determining a second score from alignment between an observed distribution of a second feature in the plurality of data points, and a reference distribution of the second feature in the semantic data type, the alignment comprising a Kolmogorov-Smirnov distance; determining a third score from alignment between an observed distribution of a third feature in the plurality of data points, and a reference distribution of the third feature in the semantic data type; calculating a total score for the semantic data type from the first, second, and third scores; determining an initial ranking of the total score in comparison with another total score reflecting the first, second, and third features in another semantic data type; identifying a fourth feature differentiating between the semantic data type and the other semantic data type; assigning a final ranking to the total score and the other total score by, determining alignment between an observed distribution of the fourth feature in the plurality of data points, and reference distributions of the fourth feature in the semantic data type and in the other semantic data type; assigning the semantic data type to the plurality of data points based upon the final ranking; and storing the plurality of data points with metadata indicating the semantic data type. 11. A non-transitory computer readable storage medium as in claim 10 wherein the semantic data type comprises an integer, a date, a month, a day, an identifier, an IP address, a currency transaction, or a currency balance. 12. A non-transitory computer readable storage medium as in claim 10 wherein the second feature comprises a data point length, a data point first digit, other than a first digit of a data point, a dispersion, a number of digits after a decimal point. 13. A non-transitory computer readable storage medium as in claim 10 wherein the method further comprises: deriving the reference distribution of the second feature from training data; receiving feedback regarding accuracy of the assignment of semantic data type; and updating the training data to reflect the feedback. 14. A non-transitory computer readable storage medium as in claim 10 wherein the method further comprises: receiving feedback regarding accuracy of the assignment of semantic data type; calculating the total score by assigning respective weight coefficients to the first, second, and third scores; and updating the respective weight coefficients based upon the feedback. 15. A non-transitory computer readable storage medium as in claim 10 wherein converting the plurality of data points to the semantic data type employs a pattern matching mechanism comprising regular expressions specifying class-specific search patterns. 16. A computer system comprising: one or more processors; a software program, executable on said computer system, the software program configured to cause an in-memory database engine to: receive a plurality of data points lacking associated metadata; determine a first score of a first feature comprising a success rate converting the plurality of data points to a semantic data type; determine a second score from alignment between an observed distribution of a second feature in the plurality of data points, and a reference distribution of the second feature in the semantic data type; determine a third score from alignment between an observed distribution of a third feature in the plurality of data points, and a reference distribution of the third feature in the semantic data type; calculate a total score for the semantic data type from the first, second, and third scores; determine an initial ranking of the total score in comparison with another total score reflecting the first, second, and third features in another semantic data type; identify a fourth feature differentiating between the semantic data type and the other semantic data type; assign a final ranking to the total score and the other total score by, determine alignment between an observed distribution of the fourth feature in the plurality of data points, and reference distributions of the fourth feature in the semantic data type and in the other semantic data type; assign the semantic data type to the plurality of data points based upon the final ranking; and store the plurality of data points in a column of a table of an in-memory database with metadata indicating the semantic data type. 17. A computer system as in claim 16 wherein the software is further configured to cause the in-memory database engine to: calculate the alignment between the observed distribution of the second feature and the reference distribution of the second feature as a Kolmogorov-Smirnov distance. 18. A computer system as in claim 16 wherein the software is further configured to cause the in-memory database engine to: receive feedback regarding accuracy of the assignment of semantic data type; calculate the total score by assigning respective weight coefficients to the first, second, and third scores; and update the respective weight coefficients based upon the feedback. 19. A computer system as in claim 16 wherein the software is further configured to cause the in-memory database engine to: derive the reference distribution of the second feature from training data stored in the in-memory database; receive feedback regarding accuracy of the assignment of semantic data type; and update the training data to reflect the feedback. 20. A computer system as in claim 16 wherein the software is further configured to cause the in-memory database engine to convert the plurality of data points to the semantic data type by pattern matching comprising regular expressions specifying class-specific search patterns. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Embodiments associate a relevant semantic data type (e.g., date) with incoming raw data (e.g., a column of digits) which lacks metadata. Assignment of semantic data type is inferred from a plurality of features. A first step determines a first feature comprising success rate in converting the raw data into various semantic data types. Then, alignment between observed/reference distributions of other features (e.g., data first digit, data length) is determined per-semantic data type. Total scores for each semantic data type are calculated from the combined features, and used as a basis for ranking the semantic data types. The total scores may reflect a weighting of the various features. In a second step, top-ranked semantic data types may be further differentiated from one another by applying additional features. User feedback regarding accuracy of semantic data type assignment, may be incorporated into training data used to modify the feature reference distributions. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Embodiments associate a relevant semantic data type (e.g., date) with incoming raw data (e.g., a column of digits) which lacks metadata. Assignment of semantic data type is inferred from a plurality of features. A first step determines a first feature comprising success rate in converting the raw data into various semantic data types. Then, alignment between observed/reference distributions of other features (e.g., data first digit, data length) is determined per-semantic data type. Total scores for each semantic data type are calculated from the combined features, and used as a basis for ranking the semantic data types. The total scores may reflect a weighting of the various features. In a second step, top-ranked semantic data types may be further differentiated from one another by applying additional features. User feedback regarding accuracy of semantic data type assignment, may be incorporated into training data used to modify the feature reference distributions. |
|
A method for updating a weight of a synapse of a neuromorphic device is provided. The synapse may include a transistor and a memristor. The memristor may have a first electrode coupled to a source electrode of the transistor. The method may include inputting a row spike to a drain electrode of the transistor at a first time; inputting a column spike to a second electrode of the memristor at a second time; inputting a row pulse to the drain electrode of the transistor at a third time that is delayed by a first delay time from the second time; inputting a column pulse to the second electrode of the memristor at a fourth time that is delayed by a second delay time from the second time; and inputting a gating pulse to a gate electrode of the transistor at a fifth time that is delayed by a third delay time from the fourth time. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for updating a weight of a synapse of a neuromorphic device, the synapse including a transistor and a memristor, the memristor having a first electrode coupled to a source electrode of the transistor, the method comprising: inputting a row spike to a drain electrode of the transistor at a first time; inputting a column spike to a second electrode of the memristor at a second time; inputting a row pulse to the drain electrode of the transistor at a third time that is delayed by a first delay time from the second time; inputting a column pulse to the second electrode of the memristor at a fourth time that is delayed by a second delay time from the second time; and inputting a gating pulse to a gate electrode of the transistor at a fifth time that is delayed by a third delay time from the fourth time. 2. The method according to claim 1, wherein the row spike is generated by a pre-synaptic neuron, and is inputted to the drain electrode of the transistor through a row line. 3. The method according to claim 1, wherein the column spike is generated by a post-synaptic neuron, and is inputted to the second electrode of the memristor through a column line. 4. The method according to claim 1, wherein the row pulse is generated by the pre-synaptic neuron, and is inputted to the drain electrode of the transistor through a row line. 5. The method according to claim 1, wherein the column pulse is generated by the post-synaptic neuron, and is inputted to the second electrode of the memristor through a column line. 6. The method according to claim 1, wherein the row spike and the column spike are generated while a gating signal is inputted to the gate electrode of the transistor. 7. The method according to claim 6, wherein the gating signal is generated by a gating controller, and is inputted to the gate electrode of the transistor through a gating line. 8. The method according to claim 1, wherein the updating of the weight of the synapse is performed for a time period in which an overlay of the row pulse and the column pulse overlaps the gating pulse. 9. A method for updating a weight of a synapse of a neuromorphic device, the method comprising: inputting, by a first neuron, a first spike to a first synapse at a first time; inputting, by a second neuron, a second spike to a second synapse at a second time that is delayed from the first time; inputting, by a third neuron, a third spike to each of the first synapse and the second synapse at a third time; inputting, by the first neuron, a first pulse to the first synapse at a fourth time that is delayed from the third time; inputting, by the second neuron, a second pulse to the second synapse at a fifth time that is delayed from the third time; inputting, by the third neuron, a third pulse to each of the first synapse and the second synapse at a sixth time; inputting a first gating pulse to the first synapse at a seventh time; and inputting a second gating pulse to the second synapse at an eighth time. 10. The method according to claim 9, wherein, if a first spike time difference from the first time to the third time is shorter than a second spike time difference from the second time to the third time, a first gating time difference from the sixth time to the seventh time is shorter than a second gating time difference from the sixth time to the eight time. 11. The method according to claim 10, wherein, if the first gating time difference is shorter than the second gating time difference, a weight of the first synapse is updated over a longer time period than a weight of the second synapse. 12. The method according to claim 9, wherein the first synapse includes a first transistor and a first memristor, the first memristor having a first electrode coupled to a source electrode of the first transistor, wherein the second synapse includes a second transistor and a second memristor, the second memristor having a first electrode coupled to a source electrode of the second transistor, wherein the first neuron is coupled to a drain electrode of the first transistor, wherein the second neuron is coupled to a drain electrode of the second transistor, and wherein the third neuron is coupled to a second electrode of the first memristor and a second electrode of the second memristor. 13. The method according to claim 12, wherein the first gating pulse is generated by a first gating controller that is coupled to a gate electrode of the first transistor, and wherein the second gating pulse is generated by a second gating controller that is coupled to a gate electrode of the second transistor. 14. The method according to claim 9, wherein the first and second neurons are first and second pre-synaptic neurons, and the third neuron is a post-synaptic neuron, and wherein the first and second synapses are coupled in common to the post-synaptic neuron. 15. The method according to claim 9, wherein the first and second neurons are first and second post-synaptic neurons, and the third neuron is a pre-synaptic neuron, and wherein the first and second synapses are coupled in common to the pre-synaptic neuron. 16. A method for updating weights of first and second synapses of a neuromorphic device, the first synapse including a first transistor and a first memristor, the first memristor having a first electrode coupled to a source electrode of the first transistor, the second synapse including a second transistor and a second memristor, the second memristor having a first electrode coupled to a source electrode of the second transistor, the method comprising: turning on the first transistor and the second transistor by inputting a first gating signal to a gate electrode of the first transistor and inputting a second gating signal to a gate electrode of the second transistor, the first gating signal being generated by a first gating controller, the second gating signal being generated by a second gating controller; inputting a first row spike to a drain electrode of the first transistor of the first synapse through a first row line, and inputting a second row spike to a drain electrode of the second transistor of the second synapse through a second row line, the first row spike being generated by a first pre-synaptic neuron coupled to the first synapse, the second row spike being generated by a second pre-synaptic neuron coupled to the second synapse; inputting a column spike generated by a post-synaptic neuron, which is coupled in common to the first synapse and the second synapse, to a second electrode of the first memristor and a second electrode of the second memristor through a column line; stopping the first and second gating signals from being input to thereby turn off the first transistor and the second transistor; inputting a first row pulse to the drain electrode of the first transistor and inputting a second row pulse to the drain electrode of the second transistor, the first row pulse being generated by the first pre-synaptic neuron, the second row pulse being generated by the second pre-synaptic neuron; inputting a column pulse to the second electrode of the first memristor and the second electrode of the second memristor, the column pulse being generated by the post-synaptic neuron; and inputting a first gating pulse to the gate electrode of the first transistor and inputting a second gating pulse to the gate electrode of the second transistor such that the first and second transistors are turned on in response to the first and second gating pulses, respectively, the first gating pulse being generated by the first gating controller, the second gating pulse being generated by the second gating controller. 17. The method according to claim 15, wherein, if a first spike time difference from a time when the first row spike is generated to a time when the column spike is generated is shorter than a second spike time difference from a time when the second row spike is generated to the time when the column spike is generated, a first gating time difference from a time when the column pulse is generated to a time when the first gating pulse is generated is shorter than a second gating time difference from the time when the column pulse is generated to a time when the second gating pulse is generated. 18. The method according to claim 17, wherein, if the first spike time difference is longer than the second spike time difference, the first gating time difference is longer than the second gating time difference. 19. The method according to claim 16, wherein the column pulse has a negative (−) voltage when the first and second row pulses have a positive (+) voltage. 20. The method according to claim 16, wherein the column pulse has a positive (+) voltage when the first and second row pulses have a negative (−) voltage. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method for updating a weight of a synapse of a neuromorphic device is provided. The synapse may include a transistor and a memristor. The memristor may have a first electrode coupled to a source electrode of the transistor. The method may include inputting a row spike to a drain electrode of the transistor at a first time; inputting a column spike to a second electrode of the memristor at a second time; inputting a row pulse to the drain electrode of the transistor at a third time that is delayed by a first delay time from the second time; inputting a column pulse to the second electrode of the memristor at a fourth time that is delayed by a second delay time from the second time; and inputting a gating pulse to a gate electrode of the transistor at a fifth time that is delayed by a third delay time from the fourth time. |
|
G06N308 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method for updating a weight of a synapse of a neuromorphic device is provided. The synapse may include a transistor and a memristor. The memristor may have a first electrode coupled to a source electrode of the transistor. The method may include inputting a row spike to a drain electrode of the transistor at a first time; inputting a column spike to a second electrode of the memristor at a second time; inputting a row pulse to the drain electrode of the transistor at a third time that is delayed by a first delay time from the second time; inputting a column pulse to the second electrode of the memristor at a fourth time that is delayed by a second delay time from the second time; and inputting a gating pulse to a gate electrode of the transistor at a fifth time that is delayed by a third delay time from the fourth time. |
|
A system and method for projecting the likely future path of the subject of a sequential decision problem. The subject of the sequential decision problem takes an action beginning with the starting state of affairs and probabilistically transitions into other states according to the structure of the decision problem, the solution to the decision problem, possibly random events, and the decisions of the subject. The likely future path consists of a sequence of actions taken by the subject, the states the subject will likely be in after taking the projected actions, and the rewards the subject is likely to receive along the future path. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A computer-aided decision making system, comprising: (a) a user input device; (b) a user output device; and (c) a processor programmed to evaluate decision problems available to a user; the programmed processor: (A) facilitating input of information from the user via the user input device, the information including (i) the decision problem to be solved (ii) an action set, the action set has elements representing actions available to the subject, each element in the action set having a corresponding action cost, the corresponding action costs forming an action cost set, (iii) at least one state dimension representing conditions relevant to the decision problem, (iv) each state dimension having a corresponding reward vector representing the reward associated with the elements of the state dimension, before consideration of the action cost set, (v) each state dimension having a corresponding transition matrix containing, for each element in the state dimension, a probability of moving from each state in the state dimension to each state in the state dimension for each action in the action set, (vi) a time index and a discount factor, the time index containing decision points available to the user, each decision point representing a point in time when the user selects from the action set, and the discount factor representing the subject's preference for rewards relative to time, (B) the programmed processor combining the reward vectors with the action cost set to form a reward matrix and the programmed processor combining the transition matrices with the action set to form a total transition matrix; (C) the programmed processor forming a functional equation from the state dimensions, the reward matrix, the total transition matrix, and all remaining inputs; (D) the programmed processor evaluating the functional equation, including error-checking and validating the inputs and performing a convergence check to ensure that the functional equation will be solvable, and the programmed processor solving the functional equation; (E) the programmed processor generating an optimal policy by using the solved functional equation to find, for every point in the decision index, the overall value-maximizing action, the set of decisions made at each point in the decision index forming the optimal policy; (F) the programmed processor generating a likely path for at least one starting state by (i) identifying an assumed action by selecting the value-maximizing action for each potential state at the initial point in the decision index, based upon the optimal policy (ii) evaluating, for the assumed action, a likely transition to occur by comparing the probabilities in the transition matrix for the combination of state dimensions; (iii) generating the likely path for each decision point in the decision index by selecting the likely transition from the possible transitions at each decision point based upon the current state, the reward in the current state given the assumed action and the most likely transition at the next decision point in the decision index (G) the programmed processor outputting the likely path to the user through the user output device. 2. A computer-aided decision making system according to claim 1, wherein the programmed processor evaluates, for the assumed actions, the most likely transition and generates the likely path for each decision point in the decision index using the most likely transition. 3. A computer-aided decision making system according to claim 1, wherein the programmed processor evaluates, for the assumed action, the most likely transition and generates the likely path for each decision point in the decision index using the most likely transition and the programmed processor generates a second likely path for a less likely transition based upon user-provided alternate path-selection criteria. 4. A computer-aided decision making system according to claim 3, wherein the programmed processor evaluates the most likely transition and at least one less likely transition and determines if the transitions are alternate paths where the likely transition probabilities, are close, or within a user-defined parameter, the programmed processor selects among alternate paths according to alternate path-selection criteria. 5. A computer-aided decision making system according to claim 1, wherein the programmed processor displays the optimal policies along the likely paths for different starting states at the initial point in the decision index. 6. A computer-aided decision making system according to claim 1, wherein the programmed processor displays the rewards along the likely paths for different starting states at the initial point in the decision index. 7. A computer-aided decision making system according to claim 1, wherein the programmed processor receives subject personality information and uses the subject personality information to determine the subject's path-selection criteria or otherwise define the assumed action. 8. A computer-aided decision making system according to claim 3, wherein the programmed processor, upon evaluating at least two alternate paths, selects the least likely transition. 9. A computer-aided decision making system according to claim 1, wherein the programmed processor prompts the user to select a subset of the state dimensions and projects a likely path for the selected state dimensions 10. A computer-aided decision making system according to claim 1, wherein the programmed processor prompts the user to select an analysis of the state dimensions in the decision problem, and the computer system conducts the analysis of dimensions to determine one or more significant state dimensions, and projects the likely path for each significant state dimension(s). 11. A computer implemented method for assisting a user in making a decision comprising: (a) providing a computer system having a user input device, a user output device, and a processor programmed with instructions to evaluate a decision problem available to the user, the instructions programming the processor: (b) using the computer system to provide the user with an option for selecting the decision problem to be solved, the user inputs information via the user input device to define the decision problem, the information including (i) the decision problem to be solved (ii) an action set, the action set has elements representing actions available to the subject, each element in the action set having a corresponding action cost, the corresponding action costs forming an action cost set, (iii) at least one state dimension representing conditions relevant to the decision problem, (iv) each state dimension having a corresponding reward vector representing the reward associated with the elements of the state dimension, before consideration of the action cost set, (v) each state dimension having a corresponding transition matrix containing, for each element in the state dimension, a probability of moving from each state in the state dimension to each state in the state dimension for each action in the action set, (vi) a time index and a discount factor, the time index containing decision points available to the user, each decision point representing a point in time when the user selects from the action set, and the discount factor representing the subject's preference for rewards relative to time, (c) forming, by the computer system manipulating the reward vectors with the action cost set, a reward matrix, and by the computer system manipulating the transition matrices with the set of actions, a total transition matrix, (d) forming, by the computer system manipulating the state dimensions, the reward matrix, the total transition matrix and all remaining inputs, a functional equation. (e) evaluating, by the computer system, the functional equation including error-checking and validating the inputs and performing a convergence check to ensure that the functional equation will be solvable, and solving the functional equation; (f) generating, by the computer system, an optimal policy by using the solved functional equation to find for every point in the decision index, the overall value-maximizing action, the set of decisions made at each point in the decision index forming the optimal policy; (g) generating, by the computer system, a likely path for at least one starting state by (i) identifying an assumed action by selecting the value-maximizing action for each potential state at the initial point in the decision index, based upon the optimal policy (ii) evaluating, for the assumed action, the most likely transition to occur by comparing the probabilities in the transition matrix for the combination of state dimensions; (iii) generating the likely path for each decision point in the decision index by selecting the most likely transition from the possible transitions at each decision point based upon the current state, the reward in the current state given the assumed action and the most likely transition at the next decision point in the decision index (E) outputting, by the computer system, the likely path to the user through the user output device. 12. A method as set forth in claim 11, wherein the computer system evaluates, for the assumed action, the most likely transition and generates the likely path for each decision point in the decision index using the most likely transition. 13. A method as set forth in claim 11, wherein the computer system evaluates, for the assumed action, the most likely transition and generates the likely path for each decision point in the decision index using the most likely transition and generates a second likely path for a less likely transition based upon user-provided alternate path-selection criteria 14. A method as set forth in claim 13, wherein the computer system evaluates the most likely transition and at least one less likely transition and determines if the transitions are alternate paths where the likely transition probabilities, are close, or within a user-defined parameter, the programmed processor selects among alternate paths according to alternate path-selection criteria. 15. A method as set forth in claim 11, wherein the computer system displays the optimal policies along the likely paths for different starting states at the initial point in the decision index. 16. A method as set forth in claim 11, wherein the computer system displays the rewards along the likely paths for different starting states at the initial point in the decision index. 17. A method as set forth in claim 11, wherein the computer system receives subject personality information and uses the subject personality information to determine the subject's path-selection criteria or otherwise define the assumed action. 18. A method as set forth in claim 13, wherein the computer system, upon evaluating at least two alternate paths, selects the least likely transition. 19. A method as set forth in claim 11 wherein the computer system prompts the user to select a subset of the state dimensions and projects a likely path for the selected state dimensions. 20. A method as set forth in claim 11 wherein the computer system prompts the user to select an analysis of the state dimensions in the decision problem, and the computer system conducts the analysis of dimensions to determine one or more significant state dimensions, and projects the likely path for each significant state dimension(s). |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A system and method for projecting the likely future path of the subject of a sequential decision problem. The subject of the sequential decision problem takes an action beginning with the starting state of affairs and probabilistically transitions into other states according to the structure of the decision problem, the solution to the decision problem, possibly random events, and the decisions of the subject. The likely future path consists of a sequence of actions taken by the subject, the states the subject will likely be in after taking the projected actions, and the rewards the subject is likely to receive along the future path. |
|
G06N5045 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A system and method for projecting the likely future path of the subject of a sequential decision problem. The subject of the sequential decision problem takes an action beginning with the starting state of affairs and probabilistically transitions into other states according to the structure of the decision problem, the solution to the decision problem, possibly random events, and the decisions of the subject. The likely future path consists of a sequence of actions taken by the subject, the states the subject will likely be in after taking the projected actions, and the rewards the subject is likely to receive along the future path. |
|
In one embodiment, a learning data processor determines a plurality of machine learning features in a computer network to collect. Upon receiving data corresponding to the plurality of features, the learning data processor may aggregate the data, and pushes the aggregated data for select features to interested learning machines associated with the computer network. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method, comprising: determining a plurality of machine learning features in a computer network to collect at a learning data processor; receiving data corresponding to the plurality of features; aggregating the data; and pushing the aggregated data for select features to interested learning machines associated with the computer network. 2. The method as in claim 1, wherein pushing comprises: multicasting each type of aggregated data within a corresponding multicast group, wherein interested learning machines join multicast groups for features in which they are interested. 3. The method as in claim 1, further comprising: receiving, from the interested learning machines, an indication of features in which they are interested. 4. The method as in claim 1, further comprising: establishing and using a control multicast group for the learning data processor to communicate control messages with all learning machines associated with the computer network. 5. The method as in claim 1, further comprising: using a control application port for the learning data processor to communicate control messages with all learning machines associated with the computer network. 6. The method as in claim 1, further comprising: selecting the selected features to push as only those collected features in which one or more learning machines associated with the network are interested. 7. The method as in claim 1, further comprising: generating a statistical analysis of all collected features; and performing a traffic reduction measure based on the statistical analysis. 8. The method as in claim 7, wherein the traffic reduction measure comprises: combining correlated features into a representative feature. 9. The method as in claim 7, wherein the traffic reduction measure comprises: tuning a rate at which updates of the aggregated data are pushed to the learning machines. 10. The method as in claim 9, further comprising: tuning the rate as a function of feature variability. 11. The method as in claim 1, further comprising: learning a probabilistic model of time evolution of the collected feature data. 12. The method as in claim 11, further comprising: extrapolating missing feature data based on the probabilistic model. 13. The method as in claim 11, further comprising: extrapolating future feature data based on the probabilistic model. 14. The method as in claim 11, further comprising: associating the predictive model with a level of confidence. 15. An apparatus, comprising: one or more network interfaces that communicate with a computer network; a processor coupled to the one or more network interfaces and configured to execute a process; and a memory configured to store process executable by the processor, the process when executed operable to: determine a plurality of machine learning features in the computer network to collect as a learning data processor; receive data corresponding to the plurality of features; aggregate the data; and push the aggregated data for select features to interested learning machines associated with the computer network. 16. The apparatus as in claim 15, wherein the process when executed to push is further operable to: multicast each type of aggregated data within a corresponding multicast group, wherein interested learning machines join multicast groups for features in which they are interested. 17. The apparatus as in claim 15, wherein the process when executed is further operable to: receive, from the interested learning machines, an indication of features in which they are interested. 18. The apparatus as in claim 15, wherein the process when executed is further operable to: generate a statistical analysis of all collected features; and perform a traffic reduction measure by combining correlated features into a representative feature based on the statistical analysis. 19. The apparatus as in claim 15, wherein the process when executed is further operable to: generate a statistical analysis of all collected features; and perform a traffic reduction measure by tuning a rate at which updates of the aggregated data are pushed to the learning machines based on the statistical analysis. 20. The apparatus as in claim 15, wherein the process when executed is further operable to: learn a probabilistic model of time evolution of the collected feature data; and extrapolate missing feature data based on the probabilistic model. 21. The apparatus as in claim 15, wherein the process when executed is further operable to: learn a probabilistic model of time evolution of the collected feature data; and extrapolate future feature data based on the probabilistic model. 22. A tangible, non-transitory, computer-readable media having software encoded thereon, the software, when executed by a processor, operable to: determine a plurality of machine learning features in the computer network to collect as a learning data processor; receive data corresponding to the plurality of features; aggregate the data; and push the aggregated data for select features to interested learning machines associated with the computer network. 23. The computer-readable media as in claim 22, wherein the software when executed is further operable to: generate a statistical analysis of all collected features; and perform a traffic reduction measure based on the statistical analysis. 24. The computer-readable media as in claim 22, wherein the software when executed is further operable to: learn a probabilistic model of time evolution of the collected feature data. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: In one embodiment, a learning data processor determines a plurality of machine learning features in a computer network to collect. Upon receiving data corresponding to the plurality of features, the learning data processor may aggregate the data, and pushes the aggregated data for select features to interested learning machines associated with the computer network. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: In one embodiment, a learning data processor determines a plurality of machine learning features in a computer network to collect. Upon receiving data corresponding to the plurality of features, the learning data processor may aggregate the data, and pushes the aggregated data for select features to interested learning machines associated with the computer network. |
|
Systems, methods and computer-accessible mediums can be provided that can determine an audience interest distribution(s) of content(s) by, for example, receiving first information related to a web behavior(s) of a user(s), determining second information related to a user interest distribution(s) of the user(s) based on the first information, and determining determine the audience interest distribution(s) of the content(s) based on the second information. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A non-transitory computer-accessible medium having stored thereon computer-executable instructions for determining at least one audience interest distribution of at least one content, wherein, when a computer hardware arrangement executes the instructions, the computer arrangement is configured to perform procedures comprising: receiving first information related to at least one behavior of at least one user; determining second information related to at least one user interest distribution of the at least one user based on the first information; and determining the at least one audience interest distribution of the at least one content based on the second information. 2. The computer-accessible medium of claim 1, wherein the computer arrangement is further configured to determine the audience interest distribution based on at least one probabilistic model of the second information. 3. The computer-accessible medium of claim 2, wherein the at least one probabilistic model includes a maximum likelihood estimator. 4. The computer-accessible medium of claim 1, wherein the at least one content includes at least one webpage. 5. The computer-accessible medium of claim 1, wherein the behavior includes a web behavior of the at least one user. 6. The computer-accessible medium of claim 5, wherein the web behavior includes substantially anonymous web behavior. 7. The computer-accessible medium of claim 1, wherein the behavior includes visits by the at least one user to at least one webpage. 8. The computer-accessible medium of claim 7, wherein the computer arrangement is further configured to determine the second information based on a plurality of topical interest categories associated with the at least one webpage. 9. The computer-accessible medium of claim 7, wherein the computer arrangement is further configured to determine the second information based on a plurality of topical interest categories associated with the second webpage. 10. The computer-accessible medium of claim 1, wherein the user interest distribution includes further information related to inherent preferences by the at least one user for at least one particular topic of interest. 11. The computer-accessible medium of claim 1, wherein the at least one user interest distribution includes a plurality of user interest distributions, and wherein the computer arrangement is further configured to determine the audience interest distribution using a weighted mean of the user interest distributions. 12. The computer-accessible medium of claim 11, wherein the weighted mean is based on an expected number of views of the at least one content. 13. The computer-accessible medium of claim 1, wherein the computer arrangement is further configured to model the at least one user interest distribution using at least one matrix, and wherein each row vector of the at least one matrix represents the at least one user's user interest distribution and each column of the at least one matrix represents a category's audience interest for all users. 14. The computer-accessible medium of claim 1, wherein the computer arrangement is further configured to determine the at least one audience interest distribution based on a multinomial distribution model. 15. The computer-accessible medium of claim 1, wherein the computer arrangement is further configured to determine the second information by inferring the at least one user interest distribution based on an inference model. 16. The computer-accessible medium of claim 15, wherein the inference model is an estimation of the at least one user's inherent interest distribution based on the at least one behavior of the at least one user. 17. The computer-accessible medium of claim 15, wherein the computer arrangement is further configured to generate the inference model by probabilistically modeling visits of the at least one user to a plurality of websites. 18. The computer-accessible medium of claim 1, wherein the computer arrangement is further configured to model the behavior of the at least one user using at least one bipartite graph. 19. The computer-accessible medium of claim 1, wherein the at least one behavior excludes information related to the at least one content. 20. A method for determining at least one audience interest distribution of at least one content, comprising: receiving first information related to at least one web behavior of at least one user; determining second information related to at least one user interest distribution of the at least one user based on the first information; and using a computer hardware arrangement, determining the at least one audience interest distribution of the at least one content based on the second information. 21-38. (canceled) 39. A system for determining at least one audience interest distribution of at least one content, comprising: a computer processing arrangement configured to: receive first information related to at least one web behavior of at least one user; determine second information related to at least one user interest distribution of the at least one user based on the first information; and determine the at least one audience interest distribution of the at least one content based on the second information. 40-60. (canceled) |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Systems, methods and computer-accessible mediums can be provided that can determine an audience interest distribution(s) of content(s) by, for example, receiving first information related to a web behavior(s) of a user(s), determining second information related to a user interest distribution(s) of the user(s) based on the first information, and determining determine the audience interest distribution(s) of the content(s) based on the second information. |
|
G06N504 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Systems, methods and computer-accessible mediums can be provided that can determine an audience interest distribution(s) of content(s) by, for example, receiving first information related to a web behavior(s) of a user(s), determining second information related to a user interest distribution(s) of the user(s) based on the first information, and determining determine the audience interest distribution(s) of the content(s) based on the second information. |
|
A method and system for validating energy measurement in a high pressure gas distribution network. The method comprises the steps of calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; measuring an actual energy value of the gas flow; and comparing the validation energy value and the actual energy value, wherein the actual energy value is validated if the validation energy value and the actual energy value are substantially equal. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for validating energy measurement in a high pressure gas distribution network, the method comprising the steps of: calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; measuring a field energy value of the gas flow; and comparing the validation energy value and the measured energy value, wherein the measured energy value is validated if the validation energy value and the measured energy value are substantially equal. 2. The method as claimed in claim 1, wherein the ANN engine is programmed to represent an energy value prediction model based on the measured parameters. 3. The method as claimed in claim 1, wherein the measured parameters comprise a gross volume, a pressure, a temperature, a specific gravity, and a calorific value of the gas flow. 4. The method as claimed in claim 1, wherein the ANN engine comprises a multilayered perceptron network structure. 5. The method as claimed in claim 1, further comprising determining a percentage difference between the validation energy value and the measured energy value. 6. The method as claimed in claim 5, further comprising identifying an alarm event based on the percentage difference exceeding a threshold. 7. The method as claimed in claim 1, comprising providing the measured parameters as a block of data comprising respective sets of the measured parameters and respective measured energy values over a selected time period; calculating respective validation energy values based on the respective sets of the measured parameters; and plotting both the measured energy values and the calculated validation energy values. 8. The method as claimed in claim 1, further comprising learning the energy value prediction model. 9. The method as claimed in claim 8, wherein the learning comprises: providing historical data for the measured parameters and the measured energy value; and applying a learning algorithm to the ANN engine based on the historical data. 10. The method as claimed in claim 9, wherein providing the historical data comprises scaling the historical data for statistical standardization. 11. A system for validating energy measurement in a high pressure gas distribution network, comprising: means for calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; means for measuring a field energy value of the gas flow; and means for comparing the validation energy value and the measured energy value, wherein the measured energy value is validated if the validation energy value and the measured energy value are substantially equal. 12. The system as claimed in claim 11, wherein the ANN engine is programmed to represent an energy value prediction model based on the measured parameters. 13. The system as claimed in claim 11, wherein the measured parameters comprise a gross volume, a pressure, a temperature, a specific gravity, and a calorific value of the gas flow. 14. The system as claimed in claim 11, wherein the ANN engine comprises a multilayered perceptron network structure. 15. The system as claimed in claim 11, further comprising means for determining a percentage difference between validation energy value and the measured energy value. 16. The system as claimed in claim 15, further comprising means for identifying an alarm event based on the percentage difference exceeding a threshold. 17. The system as claimed in claim 11, comprising: means for providing the measured parameters as a block of data comprising respective sets of the measured parameters and respective measured energy values over a selected time period; means for calculating respective validation energy values based on the respective sets of the measured parameters; and means for plotting both the measured energy values and the calculated validation energy values. 18. The system as claimed in claim 11, further comprising means for learning the energy value prediction model. 19. The system as claimed in claim 18, wherein the means for learning comprises: means for providing historical data for the measured parameters and the measured energy value; and means for applying a learning algorithm to the ANN engine based on the historical data. 20. The system as claimed in claim 19, wherein means for providing the historical data comprises means for scaling the historical data for statistical standardization. 21. A data storage medium comprising computer code for instructing a computing device to execute a method for validating energy measurement in a high pressure gas distribution network, the method comprising the steps of: calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; measuring a field energy value of the gas flow; and comparing the validation energy value and the measured energy value, wherein the measured energy value is validated if the validation energy value and the measured energy value are substantially equal. |
|
ACCEPTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: A method and system for validating energy measurement in a high pressure gas distribution network. The method comprises the steps of calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; measuring an actual energy value of the gas flow; and comparing the validation energy value and the actual energy value, wherein the actual energy value is validated if the validation energy value and the actual energy value are substantially equal. |
|
G06N302 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A method and system for validating energy measurement in a high pressure gas distribution network. The method comprises the steps of calculating a validation energy value using an artificial neural network (ANN) engine based on measured parameters associated with a gas flow in the gas distribution network; measuring an actual energy value of the gas flow; and comparing the validation energy value and the actual energy value, wherein the actual energy value is validated if the validation energy value and the actual energy value are substantially equal. |
|
A calculation device includes an adding unit configured to add at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class. The calculation device includes an accepting unit configured to accept, as input data, training data belonging to a second subclass contained in the predetermined class. The calculation device includes a calculation unit configured to calculate coupling coefficients between the new node added by the adding unit and other nodes to learn a feature of the training data belonging to the second subclass based on an output result obtained when the training data accepted by the accepting unit is input to the network. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A calculation device comprising: an adding unit configured to add at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class; an accepting unit configured to accept, as input data, training data belonging to a second subclass contained in the predetermined class; and a calculation unit configured to calculate coupling coefficients between the new node added by the adding unit and other nodes to learn a feature of the training data belonging to the second subclass based on an output result obtained when the training data accepted by the accepting unit is input to the network. 2. The calculation device according to claim 1, wherein the calculation unit calculates coupling coefficients that minimize an error between the output result obtained when the training data is input to the network and a correct output result corresponding to the training data. 3. The calculation device according to claim 2, wherein the calculation unit calculates coupling coefficients between the new node added by the adding unit and the other nodes, that minimize the error, with stabilizing coupling coefficients between the other nodes. 4. The calculation device according to claim 2, wherein the calculation unit calculates the coupling coefficients between the new node added by the adding unit and the other nodes, that minimize the error, in order from a coupling coefficient with respect to a node located closest to an output layer among nodes contained in the network. 5. The calculation device according to claim 1, wherein the adding unit adds the new node to the network by setting the coupling coefficients between nodes contained in the network and the new node to initial values that do not influence to the output result. 6. The calculation device according to claim 1, further comprising: an output unit configured to output the coupling coefficients calculated by the calculation unit as a vector indicating a feature amount of the training data. 7. The calculation device according to claim 1, further comprising: an output unit configured to output a network, to which the new node is added and in which coupling coefficients between each of nodes are set to the coupling coefficients calculated by the calculation unit. 8. A calculation method implemented by a calculation device, the calculation method comprising: adding at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class; accepting, as input data, training data belonging to a second subclass contained in the predetermined class; and calculating coupling coefficients between the new node added at the adding and other nodes to learn a feature of the training data belonging to the second subclass based on an output result obtained when the training data accepted at the accepting is input to the network. 9. A non-transitory recording medium storing a calculating program causing a computer to execute calculating process, the calculating process comprising: adding at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class; accepting, as input data, training data belonging to a second subclass contained in the predetermined class; and calculating coupling coefficients between the new node added at the adding and other nodes to learn a feature of the training data belonging to the second subclass, based on an output result obtained when the training data accepted at the accepting is input to the network. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: A calculation device includes an adding unit configured to add at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class. The calculation device includes an accepting unit configured to accept, as input data, training data belonging to a second subclass contained in the predetermined class. The calculation device includes a calculation unit configured to calculate coupling coefficients between the new node added by the adding unit and other nodes to learn a feature of the training data belonging to the second subclass based on an output result obtained when the training data accepted by the accepting unit is input to the network. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: A calculation device includes an adding unit configured to add at least one new node to a network, which has multiple nodes that output results of calculations on input data are connected and which learned a feature of data belonging to a first subclass contained in a predetermined class. The calculation device includes an accepting unit configured to accept, as input data, training data belonging to a second subclass contained in the predetermined class. The calculation device includes a calculation unit configured to calculate coupling coefficients between the new node added by the adding unit and other nodes to learn a feature of the training data belonging to the second subclass based on an output result obtained when the training data accepted by the accepting unit is input to the network. |
|
Emotional/cognitive state presentation is described. When two or more users, each using a device configured to present emotional/cognitive state data, are in proximity to one another, each device communicates an emotional/cognitive state of the user of the device to another device. Upon receiving data indicating an emotional/cognitive state of another user, an indication of the emotional/cognitive state of the user is presented. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method comprising: detecting, at a first device associated with a first user, a second device associated with a second user; determining, by the first device, an emotional or cognitive state of the first user; at least partly in response to detecting the trusted relationship, communicating to the second device, an indication of the emotional or cognitive state of the first user. 2. A method as recited in claim 1, wherein detecting the second device associated with the second user comprises detecting that the second device is in proximity to the first device. 3. A method as recited in claim 1, wherein detecting the second device associated with the second user comprises determining that a trusted relationship exists between the first user and the second user. 4. A method as recited in claim 1, wherein determining the emotional or cognitive state of the first user comprises receiving biometric sensor data indicating the emotional or cognitive state of the first user. 5. A method as recited in claim 1, further comprising: receiving from the second device, data identifying an emotional or cognitive state of the second user; presenting an indicator of the emotional or cognitive state of the second user. 6. A method as recited in claim 5, wherein presenting the indicator of the emotional or cognitive state of the second user includes presenting a virtual indicator within a virtual reality environment or mixed reality environment visible to the first user. 7. A method as recited in claim 6, wherein the virtual indicator includes a colored aura associated with the second user or a representation of the second user. 8. A system comprising: a device proximity detection module associated with a first device configured to detect a second device in proximity to the first device, wherein the first device is associated with a first user and the second device is associated with a second user; a sensor data analysis module associated with the first device, the sensor data analysis module configured to determine an emotional or cognitive state of the first user based at least in part on sensor data representing a physiological condition of the first user; and a communication interface configured to send, from the first device to the second device, data indicating the emotional or cognitive state of the first user. 9. A system as recited in claim 8, wherein the communication interface is further configured to receive at the first device, from the second device, data indicating an emotional or cognitive state of the second user. 10. A system as recited in claim 9, further comprising a presentation interface, wherein the presentation interface is configured to present the data indicating the emotional or cognitive state of the second user. 11. A system as recited in claim 10, wherein the presentation interface comprises a display device configured to display an augmented reality view. 12. A system as recited in claim 8, further comprises a biometric sensor to capture the sensor data. 13. A system as recited in claim 8, wherein the analysis module is implemented as a digital neural network (DNN) chip. 14. A system as recited in claim 8, implemented as a head-mounted display device. 15. One or more computer readable media having computer-executable instructions stored thereon, which, when executed by a first computing device, cause the first computing device to perform operations comprising: detecting a second computing device in proximity to the first computing device, wherein the first computing device is associated with a first user and the second computing device is associated with a second user; determining, based on sensor data, an emotional or cognitive state of the first user; and transmitting from the first computing device to the second device, data indicating the emotional or cognitive state of the first user. 16. One or more computer readable media as recited in claim 15, wherein detecting a second computing device in proximity to the first computing device further comprises determining a trusted relationship between the first user and the second user. 17. One or more computer readable media as recited in claim 15, the operations further comprising: receiving at the first computing device, from the second computing device, data indicating an emotional or cognitive state of the second user; presenting, for the first user, an indication of the emotional or cognitive state of the second user. 18. One or more computer readable media as recited in claim 17, wherein the indication of the emotional or cognitive state of the second user comprises a visual indication within an augmented reality view. 19. One or more computer readable media as recited in claim 15, the operations further comprising capturing the sensor data using a biometric sensor. 20. One or more computer readable media as recited in claim 15, implemented as a component of a head-mounted display device. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Emotional/cognitive state presentation is described. When two or more users, each using a device configured to present emotional/cognitive state data, are in proximity to one another, each device communicates an emotional/cognitive state of the user of the device to another device. Upon receiving data indicating an emotional/cognitive state of another user, an indication of the emotional/cognitive state of the user is presented. |
|
G06N5022 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Emotional/cognitive state presentation is described. When two or more users, each using a device configured to present emotional/cognitive state data, are in proximity to one another, each device communicates an emotional/cognitive state of the user of the device to another device. Upon receiving data indicating an emotional/cognitive state of another user, an indication of the emotional/cognitive state of the user is presented. |
|
Example systems and methods of developing a learning model are presented. In one example, a sample data set to train a first learning algorithm is accessed. A number of states for each input of the sample data set is determined A subset of the inputs is selected, and the sample data set is partitioned into a number of partitions equal to a combined number of states of the selected inputs. A second learning algorithm is created for each of the partitions, wherein each second learning algorithm receives the unselected inputs. Each of the second learning algorithms is assigned to a processor and trained using the samples of the partition corresponding to that algorithm. Decision logic is generated to direct each of a plurality of operational data units as input to one of the second learning algorithms based on states of the selected inputs of the operational data unit. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method of developing a learning model, the method comprising: accessing a sample data set to train the learning model, wherein the learning model comprises a first learning algorithm having a number of inputs and a number of outputs, and wherein each sample of the sample data set comprises a value for each of the inputs and the outputs; determining a number of states for each of the inputs based on the sample data set; selecting a subset of the inputs; partitioning the sample data set into a number of partitions equal to a combined number of states of the selected inputs, wherein each sample of a partition exhibits a state of the selected inputs corresponding to the partition; creating a second learning algorithm for each of the partitions, wherein the second learning algorithm of a corresponding partition comprises logic of the first learning algorithm in which the state of the selected inputs corresponds to the partition, and wherein the second learning algorithm is configured to receive as input those of the inputs that are not the selected inputs; assigning each of the second learning algorithms to one of a plurality of processors of a computing system; training each of the second learning algorithms on the processor assigned to the second learning algorithm using the samples of the partition corresponding to the second learning algorithm; and generating decision logic configured to direct each of a plurality of operational data units as input to one of the second learning algorithms, wherein each of the operational data units comprises a value for each of a plurality of inputs corresponding to the inputs of the sample data set, and wherein the directing of the operational data units is based on a state of the selected inputs corresponding to the operational data unit. 2. The method of claim 1, further comprising: accessing the plurality of operational data units; providing each of the operational data units to the decision logic for execution by a corresponding one of the second learning algorithms on a corresponding one of the processors; and accessing, for each of the operational data units, an output for the operational data unit from the corresponding second learning algorithm. 3. The method of claim 1, wherein: the combined number of states of the selected inputs is a greatest whole number less than or equal to a number of the plurality of processors; and the assigning of each of the second learning algorithms comprises assigning each of the second learning algorithms to a separate one of the plurality of processors. 4. The method of claim 1, wherein: the combined number of states of the selected inputs is a least whole number greater than a number of the plurality of processors of the computing system; and the assigning of each of the second learning algorithms comprises assigning a first one of the second learning algorithms and a second one of the second learning algorithms to a same one of the plurality of processors. 5. The method of claim 4, wherein: the training of each of the second learning algorithms comprises training the first one of the second learning algorithms, followed by training the second one of the second learning algorithms. 6. The method of claim 1, further comprising: determining a value distribution of each of at least some of the inputs; wherein at least one of the selected inputs has a more uniform value distribution than at least one unselected input of the plurality of inputs. 7. The method of claim 1, wherein: the first learning algorithm comprises a first artificial neural network having a first number of hidden neurons; each of the second learning algorithms comprises a second artificial neural network; and the method further comprises setting a second number of hidden neurons of at least one of the second artificial neural networks based on the first number of hidden neurons. 8. The method of claim 7, wherein the second number of hidden neurons is equal to the first number of hidden neurons based on the first number of hidden neurons being less than twice the number of inputs minus twice a number of the selected inputs. 9. The method of claim 7, wherein the second number of hidden neurons is equal to twice the number of inputs minus twice a number of the selected inputs based on the first number of hidden neurons being less than or equal to four times the number of inputs minus four times the number of selected inputs, and being greater than or equal to twice the number of inputs minus twice the number of selected inputs. 10. The method of claim 7, wherein the second number of hidden neurons is equal to the first number of hidden neurons minus twice the number of inputs, minus twice a number of the selected inputs, based on the first number of hidden neurons being greater than four times the number of inputs minus four times the number of selected inputs. 11. The method of claim 1, wherein the determining of the number of states for each of the inputs based on the sample data set comprises determining a number of possible states for each of the selected inputs of the sample data set. 12. The method of claim 2, wherein the determining of the number of states for each of the inputs based on the sample data set comprises determining a number of employed states for each of the selected inputs of the sample data set. 13. The method of claim 12, wherein: the accessing of the operational data units comprises accessing a first operational data unit that includes a first input of the selected inputs having a state that is not employed in the sample data set; and the decision logic is further configured to direct the first operational data unit as input to the second learning algorithm corresponding to one of the employed states of the first input. 14. The method of claim 12, wherein: the accessing of the operational data units comprises accessing a first operational data unit that includes a first input of the selected inputs having a state that is not employed in the sample data set; the decision logic is further configured to direct the first operational data unit as input to at least two of the second learning algorithms, wherein each of the at least two of the second learning algorithms corresponds to one of the employed states of the first input; and the method further comprises calculating a weighted average of corresponding outputs of the at least two of the second learning algorithms to produce an output for the first operational data unit. 15. The method of claim 1, wherein the first learning algorithm and each of the second learning algorithms comprises an artificial neural network. 16. The method of claim 1, wherein the first learning algorithm and each of the second learning algorithms comprises a supervised learning algorithm. 17. The method of claim 1, wherein the first learning algorithm and each of the second learning algorithms comprises an unsupervised learning algorithm. 18. A non-transitory computer-readable storage medium comprising instructions that, when executed by at least one processor of a computing system, cause the computing system to perform operations comprising: accessing a sample data set to train a learning model, wherein the learning model comprises a first learning algorithm having a number of inputs and a number of outputs, and wherein each sample of the sample data set comprises a value for each of the inputs and the outputs; determining a number of states for each of the inputs based on the sample data set; selecting a subset of the inputs; partitioning the sample data set into a number of partitions equal to a combined number of states of the selected inputs, wherein each sample of a partition exhibits a state of the selected inputs corresponding to the partition; creating a second learning algorithm for each of the partitions, wherein the second learning algorithm of a corresponding partition comprises logic of the first learning algorithm in which the state of the selected inputs corresponds to the partition, and wherein the second learning algorithm is configured to receive as input those of the inputs that are not the selected inputs; assigning each of the second learning algorithms to one of a plurality of processors of a computing system; training each of the second learning algorithms on the processor assigned to the second learning algorithm using the samples of the partition corresponding to the second learning algorithm; and generating decision logic configured to direct each of a plurality of operational data units as input to one of the second learning algorithms, wherein each of the operational data units comprises a value for each of a plurality of inputs corresponding to the inputs, and wherein the directing of the operational data units is based on the state of the selected inputs corresponding to the operational data unit. 19. A computing system comprising: a sample data set accessing module configured to access a sample data set to train a learning model, wherein the learning model comprises a first learning algorithm having a number of inputs and a number of outputs, and wherein each sample of the sample data set comprises a value for each of the inputs and the outputs; a sample data set partitioning module configured to determine a number of states for each of the inputs based on the sample data set, select a subset of the inputs, and partition the sample data set into a number of partitions equal to a combined number of states of the selected inputs, wherein each sample of a partition exhibits a state of the selected inputs corresponding to the partition; an algorithm creation module configured to create a second learning algorithm for each of the partitions, wherein the second learning algorithm of a corresponding partition comprises logic of the first learning algorithm in which the state of the selected inputs corresponds to the partition, and wherein the second learning algorithm is configured to receive as input those of the inputs that are not the selected inputs; an algorithm assignment module configured to assign each of the second learning algorithms to one of a plurality of processors of the computing system; an algorithm training module configured to train each of the second learning algorithms on the processor assigned to the second learning algorithm using the samples of the partition corresponding to the second learning algorithm; and a decision logic generation module configured to generate decision logic configured to direct each of a plurality of operational data units as input to one of the second learning algorithms, wherein each of the operational data units comprises a value for each of a plurality of inputs corresponding to the inputs of the sample data set, and wherein the directing of the operational data units is based on the state of the selected inputs corresponding to the operational data unit. 20. The system of claim 19, further comprising: an operational data unit accessing module configured to access the plurality of operational data units, and provide each of the operational data units to the decision logic for execution by a corresponding one of the second learning algorithms on a corresponding one of the processors; and an operational output accessing module configured to access, for each of the operational data units, an output for the operational data unit from the corresponding second learning algorithm. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Example systems and methods of developing a learning model are presented. In one example, a sample data set to train a first learning algorithm is accessed. A number of states for each input of the sample data set is determined A subset of the inputs is selected, and the sample data set is partitioned into a number of partitions equal to a combined number of states of the selected inputs. A second learning algorithm is created for each of the partitions, wherein each second learning algorithm receives the unselected inputs. Each of the second learning algorithms is assigned to a processor and trained using the samples of the partition corresponding to that algorithm. Decision logic is generated to direct each of a plurality of operational data units as input to one of the second learning algorithms based on states of the selected inputs of the operational data unit. |
|
G06N99005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: Example systems and methods of developing a learning model are presented. In one example, a sample data set to train a first learning algorithm is accessed. A number of states for each input of the sample data set is determined A subset of the inputs is selected, and the sample data set is partitioned into a number of partitions equal to a combined number of states of the selected inputs. A second learning algorithm is created for each of the partitions, wherein each second learning algorithm receives the unselected inputs. Each of the second learning algorithms is assigned to a processor and trained using the samples of the partition corresponding to that algorithm. Decision logic is generated to direct each of a plurality of operational data units as input to one of the second learning algorithms based on states of the selected inputs of the operational data unit. |
|
An electronic apparatus and a method of recommending content are provided. The method of recommending content of an electronic apparatus includes determining recommendation subjects for a content based on preference information of social network members in response to a preset command being input while the content is being provided, displaying a list including the recommendation subjects of the content, and recommending the content related to a selected recommendation subject in response to a selection of at least one of the recommendation subjects included in the displayed list. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method of recommending a content for an electronic apparatus, the method comprising: determining recommendation subjects for a content based on preference information of social network members in response to a preset command being input while the content is provided; displaying a list which includes the recommendation subjects of the content; and recommending the content related to a selected recommendation subject in response to at least one of the recommendation subjects included in the list being selected. 2. The method as claimed in claim 1, wherein the determining includes: requesting the preference information of the social network members from a social network server; and determining the recommendation subjects for the extracted content based on the preference information of the social network members in response to the preference information of the social network members being received from the social network server. 3. The method as claimed in claim 2, wherein the recommending of the content includes transmitting information for the selected recommendation subject and information related to the content to the social network server so that the social network server provides the information related to the content for the selected recommendation subject. 4. The method as claimed in claim 2, wherein the preference information of the social network members is directly input by the social network members or determined by the social network server based on apparatus usage information of the social network members which is stored in the social network server. 5. The method as claimed in claim 2, wherein the social network server provides the information for the content related to the selected recommendation subject in response to a pre-designated application being executed by the selected recommendation subject. 6. The method as claimed in claim 1, wherein the displaying includes displaying the list by aligning the recommendation subjects of the list in descending order of a degree of interest in the content. 7. The method as claimed in claim 1, further comprising recommending the content to an input member in response to the member who is not included in the list being individually input. 8. The method as claimed in claim 1, wherein the recommending includes recommending the content related to all the recommendation subjects included in the displayed list in response to a specific icon included in the list being selected. 9. An electronic apparatus, comprising: a communicator configured to perform communication with an external apparatus; an input configured to receive a user command; a display; and a controller configured to determine recommendation subjects for a content based on preference information of social network members in response to a preset command being input while the content is being provided, control the display in order to display a list which includes the recommendation subjects related to the content, and control the communicator to recommend the content related to a selected recommendation subject in response to at least one of the recommendation subjects included in the list being selected through the input. 10. The electronic apparatus as claimed in claim 9, wherein the communicator is configured to perform communication with a server of a social network, wherein the controller requests the preference information of the social network members from the social network server, and determines the recommendation subjects for the content extracted based on the preference information of the social network members in response to the preference information of the social network members being received through the communicator from the social network server. 11. The electronic apparatus as claimed in claim 10, wherein the controller is configured to control the communicator to transmit to the social network server information related to the selected recommendation subject and information related to the content so that the social network server provides to the selected recommendation subject the information related to the content. 12. The electronic apparatus as claimed in claim 10, wherein the preference information of the social network members is directly input by the social network members or is determined by the social network server based on apparatus usage information of the social network members which is stored in the social network server. 13. The electronic apparatus as claimed in claim 10, wherein the social network server provides the information related to the content for the selected recommendation subject in response to a pre-designated application being executed by the selected recommendation subject. 14. The electronic apparatus as claimed in claim 9, wherein the controller controls the display in order to display the displayed list by aligning the recommendation subjects of the list in descending order of a degree of interest for the content. 15. The electronic apparatus as claimed in claim 9, wherein the controller is configured to control the communicator to recommend the content to an input member in response to social network member who is not included in the list being individually input through the input. 16. The electronic apparatus as claimed in claim 9, wherein the controller is configured to control the communicator in order to recommend the content related to all the recommendation subjects included in the displayed list in response to a specific icon included in the list being selected through the input. 17. An electronic apparatus, comprising: a communicator configured to perform communication with an external apparatus; and a controller configured to determine recommendation subjects for a content based on preference information of social network members, control a display in order to display a list which includes the recommendation subjects related to the content, and control a communicator to recommend the content related to a selected recommendation subject. 18. The electronic apparatus of claim 17, further comprising an input configured to receive a user command. 19. The electronic apparatus of claim 17, further comprising a display. 20. The electronic apparatus of claim 17, wherein the controller determines recommendation subjects for a content in response to a preset command being input while the content is being provided. 21. The electronic apparatus of claim 18, wherein the controller controls the communicator to recommend the content in response to at least one of the recommendation subjects included in the list being selected through the input. 22. The electronic apparatus as claimed in claim 21, wherein the communicator is configured to perform communication with a server of a social network, wherein the controller requests the preference information of the social network members from the social network server, and determines the recommendation subjects for the content extracted based on the preference information of the social network members in response to the preference information of the social network members being received through the communicator from the social network server. 23. The electronic apparatus as claimed in claim 22, wherein the controller is configured to control the communicator to transmit to the social network server information related to the selected recommendation subject and information related to the content so that the social network server provides to the selected recommendation subject the information related to the content. |
|
REJECTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: An electronic apparatus and a method of recommending content are provided. The method of recommending content of an electronic apparatus includes determining recommendation subjects for a content based on preference information of social network members in response to a preset command being input while the content is being provided, displaying a list including the recommendation subjects of the content, and recommending the content related to a selected recommendation subject in response to a selection of at least one of the recommendation subjects included in the displayed list. |
|
G06N502 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: An electronic apparatus and a method of recommending content are provided. The method of recommending content of an electronic apparatus includes determining recommendation subjects for a content based on preference information of social network members in response to a preset command being input while the content is being provided, displaying a list including the recommendation subjects of the content, and recommending the content related to a selected recommendation subject in response to a selection of at least one of the recommendation subjects included in the displayed list. |
|
The present invention is a new method directed for detecting anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, the method comprising: training an association-map between the initial-subspaces and feature-clusters of the plurality of data-segments, the training is responsive to a fit-criterion; concatenating the initial-subspaces into cluster-subspaces, responsive to being associated to similar feature-clusters according to the association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of the data, responsive to deviation-criterion for deviation of the new data-segment from its association to one of the feature-clusters, according to the generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for the at least one anomaly. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method directed for detecting anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, said method comprising: training an association-map between said initial-subspaces and feature-clusters of said plurality of data-segments, said training is responsive to a fit-criterion; concatenating said initial-subspaces into cluster-subspaces, responsive to being associated to similar said feature-clusters according to said association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of said data, responsive to deviation-criterion for deviation of said new data-segment from its associated one of said feature-clusters, according to said generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for said at least one anomaly. 2. The method according to claim 1, wherein said data is continuous measurement-data collected from at least one sensor; and wherein said plurality of data-segments are feature-vectors extracted from plurality of sections of said data. 3. The method according to claim 2, further comprising extracting said plurality of said feature-vectors from said plurality of sections. 4. The method according to claim 3, wherein said extracting is performed by a method selected from the group consisting of: principal component analysis (PCA), independent component analysis, minimum noise fraction, random forest embedding, non-negative matrix factorization, and any combination thereof. 5. The method according to claim 1, wherein each of said plurality of data-segments is labeled with at least one context-label; and wherein said method further comprising partitioning said plurality of data-segments to said context related initial-subspaces, responsive to a predetermined similarity in their said at least one context-label. 6. The method according to claim 5, further comprising selecting said at least one context-label from the group consisting of: days of the week, midweek- or weekend-days, time of the day, light- or dark-hours, holidays, public events, weather conditions, visibility, temperature, locations, measuring scenarios, population, and any combination thereof. 7. The method according to claims 2 and 5, wherein said data is vehicle traffic measured data. 8. The method according to claim 1 or 2, further comprising clustering said feature-clusters, using an unsupervised clustering-method. 9. The method according to claim 8, wherein at least one of the following holds true: said unsupervised clustering-method is selected from the group consisting of: K-means nearest neighbor, Density-based spatial clustering of applications with noise (DBSCAN), hierarchical clustering, Gaussian mixture and any combination thereof; said deviation-criterion and said pinpointing are determined by said unsupervised clustering-method. 10. The method according to claim 8, wherein at least one of the following holds true: said clustering is incremental; said training and said concatenating are incremental. 11. The method according to claim 1 or 10, wherein said training further comprising defining at least one additional feature-cluster associated to said data-segments of at least one of said initial-subspaces, responsive to a failure of said one of said initial-subspaces to comply with said fit-criterion. 12. The method according to claim 11, further comprising repeating said training and said concatenating, responsive to said defining of said at least one additional feature-cluster. 13. The method according to claims 5 and 8, further comprising repeating said partitioning with a different said predetermined similarity and/or repeating said clustering with a different number of clusters, responsive to a failure of at least one of said initial-subspaces to comply with said fit-criterion. 14. The method according to claim 1, further comprising selecting said fit-criterion from the group consisting of: frequency threshold, average deviation threshold, statistical properties deviation threshold, dedicated matrices, Silhouette coefficients, and any combination thereof. 15. The method according to claim 1, wherein said pinpointing and said triggering are in real-time. 16. The method according to claim 1, wherein at least one of the following holds true: said deviation is distance of said new data-segment from center from its said associated one of said feature-clusters; said deviation is distance of said new data-segment from nearest data-segment in its said associated one of said feature-clusters. 17. The method according to claim 1, further comprising selecting said trigger-criterion from the group consisting of: a predetermined number of consecutive said at least one anomaly; a predetermined number of said at least one anomaly within a selected group of said data-segments; a magnitude-threshold for said deviation; and any combination thereof. 18. A computer system for detection of anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, said detection according to method steps comprising: training an association-map between said initial-subspaces and feature-clusters of said plurality of data-segments, said training is responsive to a fit-criterion; concatenating said initial-subspaces into cluster-subspaces, responsive to being associated to similar said feature-clusters according to said association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of said data, responsive to deviation-criterion for deviation of said new data-segment from its associated one of said feature-clusters, according to said generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for said at least one anomaly; wherein said computer system comprising: an interface component, configured to receive said data-segments; a feature-extractor component, configured to extract said feature-clusters; a context-identifier component, configured for partitioning of said plurality of data-segments to said context related initial-subspaces; a mapping-machine component, configured to produce and update said generalized-association-map according to said steps of training and concatenating; and an anomaly-detector, configured for said pinpointing of said at least one anomaly and for said triggering of said automatic act. 19. A non-transitory computer readable medium (CRM) that, when loaded into a memory of a computing device and executed by at least one processor of said computing device, configured to execute the steps of a computer implemented method for detecting anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, said steps comprising: training an association-map between said initial-subspaces and feature-clusters of said plurality of data-segments, said training is responsive to a fit-criterion; concatenating said initial-subspaces into cluster-subspaces, responsive to being associated to similar said feature-clusters according to said association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of said data, responsive to deviation-criterion for deviation of said new data-segment from its associated one of said feature-clusters, according to said generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for said at least one anomaly. 20. The CRM according to claim 19, wherein at least one of the following holds true: said CRM further configured to execute step of partitioning said plurality of data-segments to said context related initial-subspaces, responsive to a predetermined similarity in their said context; said CRM further configured to execute step of clustering said feature-clusters, using an unsupervised clustering-method; said data is continuous measurement-data collected from at least one sensor, and wherein said plurality of data-segments are feature-vectors extracted from plurality of sections of said data, and said CRM further configured for extracting said plurality of said feature-vectors from said plurality of sections; said CRM further configured to execute step of defining at least one additional feature-cluster associated to said data-segments of at least one of said initial-subspaces, responsive to a failure of said one of said initial-subspaces to comply with said fit-criterion; said steps of pinpointing and triggering are in real-time. |
|
REJECTED | Please predict whether this patent is acceptable.PATENT ABSTRACT: The present invention is a new method directed for detecting anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, the method comprising: training an association-map between the initial-subspaces and feature-clusters of the plurality of data-segments, the training is responsive to a fit-criterion; concatenating the initial-subspaces into cluster-subspaces, responsive to being associated to similar feature-clusters according to the association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of the data, responsive to deviation-criterion for deviation of the new data-segment from its association to one of the feature-clusters, according to the generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for the at least one anomaly. |
|
G06N5048 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The present invention is a new method directed for detecting anomalies in monitored data having plurality of data-segments partitioned to context related initial-subspaces, the method comprising: training an association-map between the initial-subspaces and feature-clusters of the plurality of data-segments, the training is responsive to a fit-criterion; concatenating the initial-subspaces into cluster-subspaces, responsive to being associated to similar feature-clusters according to the association-map, to obtain a generalized-association-map; pinpointing at least one anomaly of at least one new data-segment of the data, responsive to deviation-criterion for deviation of the new data-segment from its association to one of the feature-clusters, according to the generalized-association-map; and triggering an automatic-act responsive to a trigger-criterion for the at least one anomaly. |
|
The disclosed systems and methods include pre-calculation, per object, of object feature bin values, for identifying close matches between objects, such as text documents, that have numerous weighted features, such as specific-length word sequences. Predetermined feature weights get scaled with two or more selected adjacent scaling factors, and randomly rounded. The expanded set of weighted features of an object gets min-hashed into a predetermined number of feature bins. For each feature that qualifies to be inserted by min-hashing into a particular feature bin, and across successive feature bins, the expanded set of weighted features get min-hashed and circularly smeared into the predetermined number of feature bins. Completed pre-calculated sets of feature bin values for each scaling of the object, together with the scaling factor, are stored for use in comparing sampled features of the object with sampled features of other objects by calculating an estimated Jaccard similarity index. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method of pre-calculation, per object, of object feature bin values for identifying close matches between objects that have numerous weighted features, the method including during min-hashing of a set of weighted features into a predetermined number of feature bins, circularly smearing a feature that qualifies to be inserted by min-hashing into a particular feature bin across successive feature bins, using a processor and memory coupled to the processor, the memory holding objects, sets of weighted features of the objects, and values placed in the feature bins by min-hashing, the circularly smearing including: treating the feature bins as a circular buffer so that a first bin follows a last bin for smearing; when a min-hashed value of a particular feature qualifies to be inserted into bin (i), a qualifying min-hashed value is considered for insertion into bin (i+1) and subsequent bins, including: increasing a min-hashed value by a predetermined increment at each successive step of insertion consideration; when an increased min-hashed value is considered for insertion into the bin (i+1) and subsequent bins, inserting the increased min-hashed value when a bin under consideration is empty or holds a bin value greater than the increased min-hashed value under consideration; and continuing consideration of the bin (i+2) and subsequent bins, through the circular buffer, until the increased min-hashed value fails to qualify to be inserted into a bin under consideration; and saving a completed set of the feature bins for one or more scalings of weights of the object together with a scaling factor or sum of feature weights for use in comparing sampled features of the object with sampled features of other objects when calculating an estimated Jaccard similarity index. 2. The method of claim 1, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value or, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins. 3. The method of claim 1, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value, and, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins, inserting a new minimum fractional part value into a feature bin under consideration when the fractional part of the min-hashed value is less than a current contents of the feature bin under consideration. 4. The method of claim 1, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value and, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins, inserting a new maximum fractional part value into a feature bin under consideration when the fractional part of the min-hashed value is greater than a current contents of the feature bin under consideration. 5. The method of claim 1, further including, for a first and second object, each having the scaling factor or the sum of feature weights: determining a maximum possible similarity between the first and second object by comparing sums of the feature weights or scaled-weights of the features of the first and second object; and when the maximum possible similarity between the first and second object is less than a predetermined threshold, rejecting similarity between the first and second object without counting matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 6. The method of claim 5, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a count of matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 7. The method of claim 6, further including: selecting two or more scaling factors to apply to the set of weighted features and mapping the scaled-weights into two or more respective sets of feature bins. 8. The method of claim 7, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a comparison of the values in selected feature bins saved for the first object with the values in the feature bins saved for a second object, wherein the selected feature bins for the first and second object have matching scaling factors. 9. The method of claim 1, further including scaling weights of a weighted set such that the sum of weights is less than or equal to an oversampling factor of samples per bin times the number of bins to fill. 10. A device that provides pre-calculation, per object, of object feature bin values for identifying close matches between objects that have numerous weighted features, the device including: a processor, memory coupled to the processor, the memory holding objects, sets of weighted features of the objects, and values placed in the feature bins by min-hashing, and computer instructions loaded into the memory that, when executed, cause the processor to implement a process that includes: during min-hashing of a set of weighted features into a predetermined number of feature bins, circularly smearing a feature that qualifies to be inserted by min-hashing into a particular feature bin across successive feature bins, the circularly smearing including: treating the feature bins as a circular buffer so that a first bin follows a last bin for smearing; when a min-hashed value of a particular feature qualifies to be inserted into bin (i), a qualifying min-hashed value is considered for insertion into bin (i+1) and subsequent bins, including: increasing a min-hashed value by a predetermined increment at each successive step of insertion consideration; when an increased min-hashed value is considered for insertion into the bin (i+1) and subsequent bins, inserting the increased min-hashed value when a bin under consideration is empty or holds a bin value greater than the increased min-hashed value under consideration; and continuing consideration of the bin (i+2) and subsequent bins, through the circular buffer, until the increased min-hashed value fails to qualify to be inserted into a bin under consideration; and saving a completed set of the feature bins for one or more scalings of weights of the object together with a scaling factor or sum of feature weights for use in comparing sampled features of the object with sampled features of other objects when calculating an estimated Jaccard similarity index. 11. The device of claim 10, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value or, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins. 12. The device of claim 10, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value, and, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins, inserting a new minimum fractional part value into a feature bin under consideration when the fractional part of the min-hashed value is less than a current contents of the feature bin under consideration. 13. The device of claim 10, further including, for a first and second object, each having the scaling factor or the sum of feature weights: determining a maximum possible similarity between the first and second object by comparing sums of the feature weights or scaled-weights of the features of the first and second object; and when the maximum possible similarity between the first and second object is less than a predetermined threshold, rejecting similarity between the first and second object without counting matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 14. The device of claim 10, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a count of matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 15. The device of claim 10, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a comparison of the values in selected feature bins saved for the first object with the values in the feature bins saved for a second object, wherein the selected feature bins for the first and second object have matching scaling factors. 16. The device of claim 10, further including scaling weights of a weighted set such that the sum of weights is less than or equal to an oversampling factor of samples per bin times the number of bins to fill. 17. A tangible non-transitory computer readable storage medium that stores program instructions that, when executed, cause a computer to implement a method for pre-calculation, per object, of object feature bin values for identifying close matches between objects that have numerous weighted features, the method including: during min-hashing of a set of weighted features into a predetermined number of feature bins, circularly smearing a feature that qualifies to be inserted by min-hashing into a particular feature bin across successive feature bins, using a processor and memory coupled to the processor, the memory holding objects, sets of weighted features of the objects, and values placed in the feature bins by min-hashing, the circularly smearing including: treating the feature bins as a circular buffer so that a first bin follows a last bin for smearing; when a min-hashed value of a particular feature qualifies to be inserted into bin (i), a qualifying min-hashed value is considered for insertion into bin (i+1) and subsequent bins, including: increasing a min-hashed value by a predetermined increment at each successive step of insertion consideration; when an increased min-hashed value is considered for insertion into the bin (i+1) and subsequent bins, inserting the increased min-hashed value when a bin under consideration is empty or holds a bin value greater than the increased min-hashed value under consideration; and continuing consideration of the bin (i+2) and subsequent bins, through the circular buffer, until the increased min-hashed value fails to qualify to be inserted into a bin under consideration; and saving a completed set of the feature bins for one or more scalings of weights of the object together with a scaling factor or sum of feature weights for use in comparing sampled features of the object with sampled features of other objects when calculating an estimated Jaccard similarity index. 18. The tangible non-transitory computer readable storage medium of claim 17, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value or, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins. 19. The tangible non-transitory computer readable storage medium of claim 17, further including min-hashing by inserting min-hashed values into bins using a fractional part of a min-hashed value produced by hashing a feature value, and, in an expanded set of feature values, the feature value combined with a sequence number, into a range spanning the predetermined number of feature bins, inserting a new minimum fractional part value into a feature bin under consideration when the fractional part of the min-hashed value is less than a current contents of the feature bin under consideration. 20. The tangible non-transitory computer readable storage medium of claim 17, further including, for a first and second object, each having the scaling factor or the sum of feature weights: determining a maximum possible similarity between the first and second object by comparing sums of the feature weights or scaled-weights of the features of the first and second object; and when the maximum possible similarity between the first and second object is less than a predetermined threshold, rejecting similarity between the first and second object without counting matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 21. The tangible non-transitory computer readable storage medium of claim 17, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a count of matches of the values in the feature bins saved for the first object with the values in the feature bins saved for the second object. 22. The tangible non-transitory computer readable storage medium of claim 17, further including, for a first and second object, each having a completed set of feature bins: calculating the estimated Jaccard similarity index based on a comparison of the values in selected feature bins saved for the first object with the values in the feature bins saved for a second object, wherein the selected feature bins for the first and second object have matching scaling factors. 23. The tangible non-transitory computer readable storage medium of claim 17, further including scaling weights of a weighted set such that the sum of weights is less than or equal to an oversampling factor of samples per bin times the number of bins to fill. 24. A method of pre-calculation, per object, of object feature bin values for identifying close matches between objects that have numerous weighted features, the method including during min-hashing of an expanded set of weighted features into a predetermined number of feature bins, circularly smearing a feature that qualifies to be inserted by min-hashing into a particular feature bin across successive feature bins, using a processor and memory coupled to the processor, the memory holding objects, sets of weighted features of the objects, and values placed in the feature bins by min-hashing, for initial weighted features of an object and the predetermined number of the feature bins used to receive the values produced by the min-hashing of the weighted features, as part of calculating an estimated Jaccard similarity index among objects: scaling initial weights to produce scaled-weighted features, using scaling factors automatically selected based on a sum of the initial weights divided by a number of features in a feature set, applying two or more of the selected scaling factors to scale the initial weights for min-hashing into one set of feature bins per scaling factor, expanding the scaled-weighted features into two or more expanded sets of weighted features of the object that includes at least as many weighted features as the predetermined number of feature bins, expanding particular scaled-weighted features into multiple feature samples in proportion to respective weights of the particular scaled weighted features; the circularly smearing including: treating the feature bins as a circular buffer so that a first bin follows a last bin for smearing; when a min-hashed value of a particular feature qualifies to be inserted into bin (i), a qualifying min-hashed value is considered for insertion into bin (i+1) and subsequent bins, including: increasing a min-hashed value by a predetermined increment at each successive step of insertion consideration; when an increased min-hashed value is considered for insertion into the bin (i+1) and subsequent bins, inserting the increased min-hashed value when a bin under consideration is empty or holds a bin value greater than the increased min-hashed value under consideration; and continuing consideration of the bin (i+2) and subsequent bins, through the circular buffer, until the increased min-hashed value fails to qualify to be inserted into a bin under consideration; and saving completed sets of the feature bins for each scaling of the object together with a scaling factor or sum of the initial weights for use in comparing sampled features of the object with sampled features of other objects by calculating an estimated Jaccard similarity index. 25. A tangible non-transitory computer readable storage medium with instructions that are combinable with a processor and memory coupled to the processor to implement the method of claim 24. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: The disclosed systems and methods include pre-calculation, per object, of object feature bin values, for identifying close matches between objects, such as text documents, that have numerous weighted features, such as specific-length word sequences. Predetermined feature weights get scaled with two or more selected adjacent scaling factors, and randomly rounded. The expanded set of weighted features of an object gets min-hashed into a predetermined number of feature bins. For each feature that qualifies to be inserted by min-hashing into a particular feature bin, and across successive feature bins, the expanded set of weighted features get min-hashed and circularly smeared into the predetermined number of feature bins. Completed pre-calculated sets of feature bin values for each scaling of the object, together with the scaling factor, are stored for use in comparing sampled features of the object with sampled features of other objects by calculating an estimated Jaccard similarity index. |
|
G06N7005 | Please help me predict the CPC LABEL for this patentPATENT ABSTRACT: The disclosed systems and methods include pre-calculation, per object, of object feature bin values, for identifying close matches between objects, such as text documents, that have numerous weighted features, such as specific-length word sequences. Predetermined feature weights get scaled with two or more selected adjacent scaling factors, and randomly rounded. The expanded set of weighted features of an object gets min-hashed into a predetermined number of feature bins. For each feature that qualifies to be inserted by min-hashing into a particular feature bin, and across successive feature bins, the expanded set of weighted features get min-hashed and circularly smeared into the predetermined number of feature bins. Completed pre-calculated sets of feature bin values for each scaling of the object, together with the scaling factor, are stored for use in comparing sampled features of the object with sampled features of other objects by calculating an estimated Jaccard similarity index. |
|
Methods, devices, and servers for friend recommendation are provided. A user association set of a target user is obtained. Original data of each associated user in the user association set is obtained. The original data include location relationship data, associated friend data, time relationship data, or combinations thereof, between each associated user and the target user. The original data of each associated user is screened to obtain feature data to form a feature collection for each associated user. A pre-configured N-Tree prediction model is used to process the feature collection for a prediction calculation to obtain an association-predicting value for each associated user. According to the association-predicting value of each associated user, a friend user for the target user from the user association set is determined and recommended to the target user. | Please help me write a proper abstract based on the patent claims. CLAIM: 1. A method for friend recommendation, comprising: obtaining a user association set of a target user, wherein the user association set comprises a plurality of associated users that are associated with the target user; obtaining original data of each associated user in the user association set of the target user, wherein the original data comprise location relationship data, associated friend data, time relationship data, or combinations thereof, between each associated user and the target user; screening the original data of each associated user to obtain feature data to form a feature collection for each associated user; using a pre-configured N-Tree prediction model to process the feature collection for a prediction calculation to obtain an association-predicting value for each associated user; and according to the association-predicting value of each associated user, determining a friend user for the target user from the user association set and recommending the friend user to the target user. 2. The method according to claim 1, wherein the N-Tree prediction model comprises a prediction model based on gradient boosting decision tree (GBDT). 3. The method according to claim 1, wherein the step of using the pre-configured N-Tree prediction model to obtain the association-predicting value for each associated user comprises: (a) determining feature collections of a plurality of predicted users; (b) configuring an initial weight value in the pre-configured N-Tree prediction model to respectively process the feature collection of each predicted user for the prediction calculation to determine the association-predicting value corresponding to the feature collection of each predicted user; (c) ranking each predicted user according to an amount of the association-predicting value to form a ranking list; (d) when a same-ranking rate between the obtained ranking list and a pre-set ranking list for the predicted users reaches a threshold value, outputting the N-Tree prediction model configured with the initial weight value as the pre-configured N-Tree prediction model; and (e) when the same-ranking rate between the obtained ranking list and the pre-set ranking list for the predicted users does not reach the threshold value, adjusting the weight value, and repeating steps (b)-(e) by: configuring the adjusted weight value in the pre-configured N-Tree prediction model to respectively process the feature collection of each predicted user for the prediction calculation to determine the association-predicting value corresponding to the feature collection of each predicted user, until the same-ranking rate between the obtained ranking list and the pre-set ranking list for the predicted users reaches the threshold value, and outputting the N-Tree prediction model configured with a finally-adjusted weight value as the pre-configured N-Tree prediction model. 4. The method according to claim 3, wherein the step of according to the association-predicting value of each associated user, determining a friend user for the target user from the user association set and recommending the friend user to the target user comprises: determining an associated user having the association-predicting value greater than a pre-set prediction threshold value as the friend user of the target user; and depending on an amount of the association-predicting value of each determined friend user, ranking the determined friend users to provide a ranking result, and recommending one or more determined friend users to the target user according to the ranking result. 5. The method according to claim 4, wherein the step of obtaining the user association set of the target user comprises: extracting marking information of the target user, and determining the plurality of associated users in the user association set according to the marking information, wherein the marking information comprises account information of the target user, location information of the target user, and combinations thereof. 6. A non-transitory computer-readable storage medium comprising instructions stored thereon, wherein, when being executed, the instructions cause one or more processors of a device to perform the method according to claim 1. 7. A device for friend recommendation, comprising: an obtaining module, configured to obtain a user association set of a target user, wherein the user association set comprises a plurality of associated users that are associated with the target user, wherein the obtaining module is further configured to obtain original data of each associated user in the user association set of the target user, wherein the original data comprise location relationship data, associated friend data, time relationship data, or combinations thereof, between each associated user and the target user; a collecting module, configured to screen the original data of each associated user to obtain feature data to form a feature collection for each associated user; a processing module, configured to use a pre-configured N-Tree prediction model to process the feature collection for a prediction calculation to obtain an association-predicting value for each associated user; and a recommending module configured, according to the association-predicting value of each associated user, to determine a friend user for the target user from the user association set and recommend the friend user to the target user. 8. The device according to claim 7, further comprising: a pre-configuring module, configured to pre-configure the N-Tree prediction model comprising a prediction model based on gradient boosting decision tree (GBDT). 9. The device according to claim 8, wherein the pre-configuring module comprises: a collection-determining unit, configured to determine feature collections of the plurality of predicted users; a first calculating unit, configured to respectively process the feature collection of each predicted user for the prediction calculation using the pre-configured N-Tree prediction model having an initial weight value to determine the association-predicting value corresponding to the feature collection of each predicted user; a ranking unit, configured to rank each predicted user according to an amount of the association-predicting value to form a ranking list; a first outputting unit configured, when a same-ranking rate between the obtained ranking list and a pre-set ranking list for the predicted users reaches a threshold value, to output the N-Tree prediction model configured with the initial weight value as the pre-configured N-Tree prediction model; a second calculating unit configured: when the same-ranking rate between the obtained ranking list and the pre-set ranking list for the predicted users does not reach the threshold value, to adjust the weight value, and to use the pre-configured N-Tree prediction model having the adjusted weight value to respectively process the feature collection of each predicted user for the prediction calculation to determine the association-predicting value corresponding to the feature collection of each predicted user, and to repeat adjustment of the weight value until the same-ranking rate between the obtained ranking list and the pre-set ranking list for the predicted users reaches the threshold value; and a second outputting module, configured to output the N-Tree prediction model having a finally-adjusted weight value as the pre-configured N-Tree prediction model. 10. The device according to claim 9, wherein the recommending module comprises: a friend determining unit, configured to determine an associated user having the association-predicting value greater than a pre-set prediction threshold value as the friend user of the target user; and a recommending unit configured, depending on an amount of the association-predicting value of each determined friend user, to rank the determined friend users to provide a ranking result, and to recommend one or more determined friend users to the target user according to the ranking result. 11. The device according to claim 10, further comprising: a determining unit, configured to extract marking information of the target user, and to determine the plurality of associated users in the user association set according to the marking information, wherein the marking information comprises account information of the target user, location information of the target user, and combinations thereof. 12. A server comprising the device according to claim 7. |
|
PENDING | Please predict whether this patent is acceptable.PATENT ABSTRACT: Methods, devices, and servers for friend recommendation are provided. A user association set of a target user is obtained. Original data of each associated user in the user association set is obtained. The original data include location relationship data, associated friend data, time relationship data, or combinations thereof, between each associated user and the target user. The original data of each associated user is screened to obtain feature data to form a feature collection for each associated user. A pre-configured N-Tree prediction model is used to process the feature collection for a prediction calculation to obtain an association-predicting value for each associated user. According to the association-predicting value of each associated user, a friend user for the target user from the user association set is determined and recommended to the target user. |
Subsets and Splits