etic-experiments

This repository contains the code to reproduce the experiments in this paper at AISTATS 2022. The paper introduces the entropy regularized optimal transport independence criterion and apply it to test for independence.

Dependencies

The code is written in Python and the dependencies are:

• python >= 3.6
• ipykernel >= 6.4.1
• matplotlib >= 3.5.0
• numpy >= 1.19.1
• pathos >= 0.2.8
• pot >= 0.6.0
• scikit-learn >= 0.22.1
• scipy >= 1.6.2
• seaborn >= 0.11.2

Conda Environment: We recommend using a conda environment. To setup the environment, run

conda env create --file environment.yml
# activate the environment
conda activate etic
python -m ipykernel install --user --name etic

Files

• ind_tests.py: code implementing independence tests.
• utils.py and file_utils.py: utility functions.
• experiments.ipynb: Jupyter notebook to run experiments and produce plots.

Citation

If you find this repository useful, or you use it in your research, please cite:

@inproceedings{liu2022entropy,
  title={{Entropy Regularized Optimal Transport Independence Criterion}},
  author={Liu, Lang and Pal, Soumik and Harchaoui, Zaid},
  booktitle={AISTATS},
  year={2022}
}

Acknowledgements

This work was supported by NSF DMS-2023166, NSF CCF-2019844, NSF DMS-2052239, PIMS CRG (PIHOT), NSF DMS-2134012, CIFAR-LMB, and faculty research awards.