Create Invention_Inspiration_and_Conception_Record

Invention_Inspiration_and_Conception_Record

@@ -0,0 +1,529 @@
+Invention Inspiration and Conception Record and Inventorship indicia.
+March 13, 2025
+https://streamyard.com/watch/XAsswGvrcyHJ
+
+StreamYard On-Air
+Time Series Mastery: Hands-On Workshops
+123 watching now
+
+​Time series forecasting is more than just predicting future trends - it’s a critical skill for industries ranging from finance to healthcare, retail, and beyond. Join us for a one-day virtual event packed with expert-led workshops designed to equip you with the latest AI-driven and classical forecasting techniques.
+
+What’s on the agenda?
+
+12.00 pm ET - Talk - Optimizing Forecast Stability and Accuracy with Genetic Algorithms by Jeff Tackes and Hamed Alikhani PhD - 30 min
+
+​12.35 pm ET - Talk - State of Foundation Models For Time Series Forecasting by Marco Peixeiro - 30 min
+
+​1.10 pm ET - Training -  Unlocking the Future with AI-Driven Time Series Forecasting by Jeffrey Yau - 2 h
+
+​3.15 pm ET - Workshop - Forecasting the Future Using Time Series by John Mount - 1 h
+
+TT
+
+Timothy Tadeo11:55 AM
+Hi All Location: Massachusetts Company: Rapid-Scale Consulting(Independent Contractor) LinkedIn: https://www.linkedin.com/in/timtadeo/
+
+
+
+👍
+4
+
+AO
+
+Anastasia ODSC11:58 AM
+Hellooo everyone! We will start in just a few minutes!
+
+
+👍
+10
+
+👋
+6
+
+PR
+
+Priyanshu Rawat11:59 AM
+Hello , i wanted to inquire if this session will be recorded
+
+MT
+
+Martial Terran11:59 AM
+Hi From New York. https://huggingface.co/MartialTerran
+
+BC
+
+Bill Chung11:59 AM
+is it also going to be recorded?
+
+CV
+
+Casper Van Coesant11:59 AM
+Hi all, tuning in form Boston, MA.
+
+HG
+
+Hanne Grosemans12:00 PM
+Hey
+
+AJ
+
+allan johns12:00 PM
+hello
+
+AC
+
+Adam Chan12:00 PM
+Yo !
+
+SM
+
+sabyasachi mukhopadhyay12:01 PM
+From Bangalore India
+
+SM
+
+Sheamus McGovern12:01 PM
+Hi from Cambridge!
+
+
+👋
+1
+
+NL
+
+Nghi Le12:01 PM
+Hi, tuning in from Indiana!
+
+SM
+
+Shawn McNulty12:01 PM
+Hello from DC
+
+MA
+
+Michelle Acevedo12:02 PM
+Hi from Chicago
+
+NM
+
+Neal Makowski12:02 PM
+hello from, boston
+
+AR
+
+AJ Ruiz12:02 PM
+Cheers from Maryland
+
+RT
+
+Ria Thazhe Punathil12:02 PM
+Hi from California
+
+DP
+
+David Pinedo12:02 PM
+Hello from New York City
+
+GR
+
+Gaylyn Ruvere12:02 PM
+Hi from Connecticut
+
+Dİ
+
+DS İlker12:02 PM
+Hi from Berlin
+
+SL
+
+Sarah Lundrogan12:02 PM
+Hello from Boston
+
+SM
+
+sabyasachi mukhopadhyay12:02 PM
+Will this sessions be recorded?
+
+LC
+
+Louis Cancino12:02 PM
+Howdy from Houston
+
+JO
+
+Joel Otolorin12:02 PM
+Hello from UK
+
+AA
+
+Anizio Anizio12:02 PM
+Hi from toronto
+
+MP
+
+Mitch Pirlot12:02 PM
+Hello from Chicago
+
+RK
+
+Ryan Kivela12:03 PM
+John Ryan Kivela Ventura, CA
+
+TB
+
+Tatiana Burkham12:03 PM
+Hello from Reno
+
+YK
+
+Yash Karle12:04 PM
+Hello from Dublin, Ireland
+
+EG
+
+Elena Gorczyca12:04 PM
+Hello from Boston
+
+JA
+
+james acosta12:04 PM
+Hi from Plantation FL
+
+TC
+
+Tom Cal12:04 PM
+Hello from Stamford, CT. https://www.linkedin.com/in/thomasical/
+
+
+👍
+1
+
+PM
+
+Paula Micetic12:06 PM
+Hi from Boston
+
+CZ
+
+Chris Zeiders12:07 PM
+Hi from Pittsburgh
+
+AP
+
+Aishwarya Patil12:07 PM
+Hello from Zurich
+
+
+👋
+1
+
+SR
+
+Sidharth Ramesh12:07 PM
+Hello from Zürich
+
+NO
+
+Nnamdi Odozi12:09 PM
+what do you mean by skew?
+
+HN
+
+Hiyab Negga12:10 PM
+Hello from Milan
+
+AA
+
+Aditya Ambardekar12:12 PM
+hi from Boston
+
+SC
+
+stanley chong12:16 PM
+from a Starbucks in Frankfurt Flughafen
+
+AL
+
+Abiola Lawani12:16 PM
+will there be recording
+
+JW
+
+Jacob Woodby12:17 PM
+Hello from NW Florida.
+
+CN
+
+Christian NGNIE12:22 PM
+is there a max number of Models you need to hit before additive aggregation and optimization the weights using GA? how does the chosen error metric affect the results?
+
+MT
+
+Martial Terran12:24 PM
+Based on the graphs (delta blue to red) it seems your models are not receiving enough information sufficient to make predictions. Why not focus in increasing the types and amounts of data (e.g., using GPT Transformer architecture for time series prediction)?
+
+MT
+
+Martial Terran12:27 PM
+Skew = SKU ?
+
+
+👍
+1
+
+CV
+
+Casper Van Coesant12:27 PM
+SKU = Stock Keeping Unit (unique ID)
+
+H
+
+Hamed12:28 PM
+Yes. Skew we meant SKU
+
+KS
+
+Kateryna Speck12:28 PM
+How do you disaggregate predictions from NSGA-III? For instance how do you split Category into SKUs?
+
+VY
+
+Vijay Yadav12:31 PM
+While preparing the forecast models, we need to determine which APIs and datasets we need to fetch and include in that model.
+
+MT
+
+MH Tan12:32 PM
+What are the limitations for integrating Genetic Algorithms for your weights optimizations?
+
+MT
+
+Martial Terran12:37 PM
+Why not pay/bribe your wholesale buyers to give you real time purchaser data, and discount based on them adhering to stable purchase schedules (3 week or one week)
+
+AO
+
+Anastasia ODSC12:37 PM
+Link to the survey - https://forms.gle/SZm8PpxzfLzYtRKH8
+
+TK
+
+Theodore Koussoulis12:38 PM
+will you record the webinar ?
+
+JT
+
+Jeff Tackes12:39 PM
+we do getsome of our larger customer "order" forecasts, and that is considered in our forecasts as well - but that also assumes our customers are good at forecasting when the BUYER will actually make the purchase.. and there is another layer of bulk buying, discounts, buyer unique impacts, etc. Which all create noise.
+
+AO
+
+Anastasia ODSC12:40 PM
+yes yes
+
+
+👍
+1
+
+DA
+
+David Arenas12:40 PM
+yes
+
+NM
+
+Neal Makowski12:40 PM
+yes and yes
+
+KM
+
+kim monzon12:40 PM
+yep
+
+AR
+
+AJ Ruiz12:40 PM
+yes and yes
+
+TB
+
+Tatiana Burkham12:40 PM
+yes
+
+NL
+
+Nghi Le12:40 PM
+Yes
+
+CN
+
+Christian NGNIE12:40 PM
+yes
+
+MT
+
+Martial Terran12:41 PM
+As you said, ever player in the chain is making predictions. It seems you all (manufacturer, wholesaler, retailer) need to subscribe to THE SAME PREDICTIVE MODEL and agree, by contract, to place orders according to ITS PREDICTIONS. The money you manufacturer save can be the incentive to participate.
+
+MA
+
+mario anthonny seminario bravo12:43 PM
+prediction or estimate?
+
+JT
+
+Jeff Tackes12:50 PM
+Martial Terran - that would be ideal - but thats not how it works in practice. With very real financial impacts on both sides and especially working with both large and small companies,telling your customer or manufacturer this is what you have to order or what you said you would order 3 months is what you HAVE to buy.. and not having the flexibility to change.. isnt going to fly.
+
+SM
+
+Sheamus McGovern12:57 PM
+https://huggingface.co/Salesforce/moirai-1.0-R-large
+
+SM
+
+Sheamus McGovern12:58 PM
+https://github.com/google-research/timesfm
+
+SM
+
+Sheamus McGovern12:58 PM
+https://docs.nixtla.io/
+
+SM
+
+Sheamus McGovern12:59 PM
+https://docs.nixtla.io/docs/getting-started-about_timegpt
+
+AO
+
+Anastasia ODSC1:00 PM
+You can see the doc to Nixtla technology here: https://nixtlaverse.nixtla.io/
+
+SM
+
+Sheamus McGovern1:03 PM
+Can you discuss again how to determine or prevent hallucinations
+
+TC
+
+Tom Cal1:04 PM
+The report "The State of Machine Learning Competitions 2024 Edition" states: "Time Series & Tabular Data ... While deep learning has clearly proved its use for tabular data, one notable omission among winning approaches in 2024 is the use of any tabular or time-series pre-trained foundation models." Q. What are your thoughts about why this is so? Q. Given this, what are the best use cases for time-series foundation models?
+
+TC
+
+Tom Cal1:04 PM
+Source: https://mlcontests.com/state-of-machine-learning-competitions-2024/
+
+CR
+
+Choudhury Rahman1:04 PM
+How accurately these stock models perform when there are clear seasonality trend and data is very sparse
+
+NO
+
+Nnamdi Odozi1:05 PM
+Are the forecasts Interpretable or Explainable eg SHAP?
+
+MT
+
+MH Tan1:06 PM
+Can you finetune on a scaled down version of those time series foundation models like using methods such as LoRA?
+
+MT
+
+Martial Terran1:06 PM
+Jeff, thank you for the acknowledgement and the insightful feedback. But, soon, all business decisions will be made by AI so human arbitrariness (supply purchase manager going on vacation, or got a divorce) can be reduced or eliminated. I am going to publish a first draft of an Transformer-based Enhanced Business Model for Collaborative Predictive Supply Chain Enhanced_Business_Model_for_Collaborative_Predictive_Supply_Chain_model.py at https://huggingface.co/MartialTerran
+
+TC
+
+Tom Cal1:07 PM
+Thanks Marco!
+
+NO
+
+Nnamdi Odozi1:10 PM
+How did the Large Time Model compare with deep neural nets eg bidirectional LSTMs?
+
+MT
+
+MH Tan1:10 PM
+For those foundation models that support features, how do you determine what type of features does the model support?
+
+MB
+
+Mohamed Boosiri1:12 PM
+How foundational model handles data structural breaks in the data? like, a huge change or break in the data?
+
+M
+
+Marco1:14 PM
+Keep the conversation going on LinkedIn: https://www.linkedin.com/in/marco-peixeiro/
+
+M
+
+Marco1:14 PM
+https://www.manning.com/books/time-series-forecasting-using-foundation-models?utm_source=marcopeix&utm_medium=affiliate&utm_campaign=book_peixeiro2&a_aid=marcopeix&a_bid=db55279d
+
+
+🙌
+2
+
+AO
+
+Anastasia ODSC1:15 PM
+Link to the survey - https://forms.gle/SZm8PpxzfLzYtRKH8
+
+EO
+
+Esther Osikoya1:16 PM
+👍
+
+NP
+
+Neel Pujara1:16 PM
+Thank you MArco!
+
+SK
+
+Sashi Kumar Nagulakonda1:16 PM
+Thank you, Marco. Great session.
+
+CN
+
+Christian NGNIE1:17 PM
+Great Marco!
+
+AO
+
+Anastasia ODSC1:19 PM
+Prerequisites: https://docs.google.com/spreadsheets/d/1DsSns9O-2mq3knCEZRv_cnkd5xJFz6uf1aG9DjbS_Cs/edit?gid=600580479#gid=600580479
+
+
+👍
+2
+
+MA
+
+mario anthonny seminario bravo1:27 PM
+the time range for viz
+
+JL
+
+Jessie Lee2:08 PM
+how will we get the noteboooks and the recordings after the session?
+
+ML
+
+Melody Lim2:09 PM
+how will the slides be shared?
+
+MT
+
+Martial Terran2:09 PM
+I have completed my disclosure of a proposed an industry-specific Collaborative Prediction Model.py and Transaction Scheduler Contract system. https://huggingface.co/MartialTerran/Contract-Enforced_Collaborative_Supply_Chain_Forecasting_Model.py
+
+MT
+
+Martial Terran2:09 PM
+This Business Model, patentable under US Patent Law within One Year of this Publication. To collaborate further or if your company may desire to obtain a Patent (by assignment from the inventor) or a License for this invention in your industry, contact me at [email protected]