Fair Representation: Guaranteeing Approximate Multiple Group Fairness for Unknown Tasks

TL;DR

This paper investigates how fair representations can ensure approximate fairness across multiple unknown prediction tasks and fairness notions, demonstrating theoretical guarantees and practical learning methods.

Contribution

It introduces a theoretical framework showing fair representations can guarantee fairness for discriminative tasks and proposes a self-supervised learning method to achieve fair, discriminative representations.

Findings

Fair representations linearly control seven group fairness notions.

Discriminative representations approximately satisfy multiple fairness notions.

Experiments confirm learned representations improve fairness in downstream tasks.

Abstract

Motivated by scenarios where data is used for diverse prediction tasks, we study whether fair representation can be used to guarantee fairness for unknown tasks and for multiple fairness notions simultaneously. We consider seven group fairness notions that cover the concepts of independence, separation, and calibration. Against the backdrop of the fairness impossibility results, we explore approximate fairness. We prove that, although fair representation might not guarantee fairness for all prediction tasks, it does guarantee fairness for an important subset of tasks -- the tasks for which the representation is discriminative. Specifically, all seven group fairness notions are linearly controlled by fairness and discriminativeness of the representation. When an incompatibility exists between different fairness notions, fair and discriminative representation hits the sweet spot that approximately satisfies all notions. Motivated by our theoretical findings, we propose to learn both fair and discriminative representations using pretext loss which self-supervises learning, and Maximum Mean Discrepancy as a fair regularizer. Experiments on tabular, image, and face datasets show that using the learned representation, downstream predictions that we are unaware of when learning the representation indeed become fairer for seven group fairness notions, and the fairness guarantees computed from our theoretical results are all valid.