src/aif360/data/raw/adult/README.md

Adult/ Census Income Data

Source / Data Set Description:

https://archive.ics.uci.edu/ml/datasets/adult

This dataset is used to predict whether income exceeds $50K/yr. Contains 48842 instances, mix of continuous and discrete. adult.data contains the first 32561 instances which comprise the original training set while adult.test contains an additional 16281 instances which comprise the original test set.

Information about dataset features is provided in adult.names

Download instructions

Download the following files and place them as-is in the current folder

1. adult.data

2. adult.test

3. adult.names

Relevant Papers

• Ron Kohavi, "Scaling Up the Accuracy of Naive-Bayes Classifiers: a Decision-Tree Hybrid", Proceedings of the Second International Conference on Knowledge Discovery and Data Mining, 1996