datasets
A collection of public datasets for supervised machine learning research. The conventions with the datasets are as follows:
- All datasets are in CSV format.
- All datasets have header rows.
- The target variable is always the last column.
- All numeric nominal features have been encoded as strings.
- Any constant columns have been removed.
- Any row ID-like columns have been removed.
- Watch out for any possible missing values in the descriptive features.
A sample Python script named "prepare_dataset_for_modeling_github.py" has also been included for loading these datasets and preparing them for model fitting.
####################################################
Description of these datasets can be found in the "github_datasets_desc" Notebook file:
https://github.com/vaksakalli/datasets/blob/master/github_datasets_desc.ipynb