A python implementation of Samadi et al's Fair PCA paper (Samadi, Samira, et al. "The price of fair pca: One extra dimension." Advances in Neural Information Processing Systems. 2018).
We evaluate the effectiveness of Fair PCA in terms of equalized odds.
Download the UCI Credit Default dataset from here and place it in the dataset folder. Install python dependencies:
pip install -r requirements.txt
The results of these experiments have been used in our publication "On Fair Representation in Machine Learning" (Akbari, 2020). The pre-print version is available here.