In this project, we deal with binary classification with LS-SVM in a large-dimensional setting. We want to do a performance analysis similar to this article, but while adding fairness constraints. We will focus on the separation property.

Clone the environment using conda

conda env create -n fair_ls_svm -f environment.yml

then

conda activate fair_ls_svm

Run the file

python run.py

To modify the parameters of the experiments, edit the corresponding option in the file run.py.