Fair Representation Learning with Controllable High Confidence Guarantees via Adversarial Inference

TL;DR

This paper introduces a novel framework called FRG that guarantees high-confidence fairness in learned representations, ensuring demographic disparities stay within user-defined thresholds across various downstream tasks.

Contribution

The paper proposes the FRG framework, which provides controllable high-confidence fairness guarantees in representation learning using an adversarial approach, a novel contribution in fairness assurance.

Findings

FRG consistently bounds unfairness across multiple downstream models.

FRG outperforms six state-of-the-art fair representation methods in empirical evaluations.

FRG effectively maintains fairness guarantees with high confidence in real-world datasets.

Abstract

Representation learning is increasingly applied to generate representations that generalize well across multiple downstream tasks. Ensuring fairness guarantees in representation learning is crucial to prevent unfairness toward specific demographic groups in downstream tasks. In this work, we formally introduce the task of learning representations that achieve high-confidence fairness. We aim to guarantee that demographic disparity in every downstream prediction remains bounded by a *user-defined* error threshold $\epsilon$, with *controllable* high probability. To this end, we propose the ***F**air **R**epresentation learning with high-confidence **G**uarantees (FRG)* framework, which provides these high-confidence fairness guarantees by leveraging an optimized adversarial model. We empirically evaluate FRG on three real-world datasets, comparing its performance to six state-of-the-art fair representation learning methods. Our results demonstrate that FRG consistently bounds unfairness across a range of downstream models and tasks.