Wasserstein Fair Classification

TL;DR

This paper introduces a Wasserstein distance-based method for fair classification that ensures independence from sensitive attributes, offering theoretical robustness and practical efficiency, with strong empirical results on benchmark datasets.

Contribution

It presents a novel fairness approach using Wasserstein-1 distances, with methods for test-time privacy and efficient implementation, advancing fair classification techniques.

Findings

Effective fairness enforcement demonstrated on benchmark datasets

Robustness to threshold choices in classification

Competitive performance against existing fairness baselines

Abstract

We propose an approach to fair classification that enforces independence between the classifier outputs and sensitive information by minimizing Wasserstein-1 distances. The approach has desirable theoretical properties and is robust to specific choices of the threshold used to obtain class predictions from model outputs. We introduce different methods that enable hiding sensitive information at test time or have a simple and fast implementation. We show empirical performance against different fairness baselines on several benchmark fairness datasets.