SLIDE: a surrogate fairness constraint to ensure fairness consistency

TL;DR

This paper introduces SLIDE, a new surrogate fairness constraint for AI models that ensures fairness consistency while being computationally feasible and asymptotically valid, with strong empirical performance.

Contribution

The paper proposes SLIDE, a novel surrogate fairness constraint that guarantees fairness consistency and fast convergence, improving upon existing methods.

Findings

SLIDE is computationally feasible for classification tasks.

SLIDE ensures asymptotic fairness constraint satisfaction.

Numerical experiments show SLIDE performs well on benchmark datasets.

Abstract

As they have a vital effect on social decision makings, AI algorithms should be not only accurate and but also fair. Among various algorithms for fairness AI, learning a prediction model by minimizing the empirical risk (e.g., cross-entropy) subject to a given fairness constraint has received much attention. To avoid computational difficulty, however, a given fairness constraint is replaced by a surrogate fairness constraint as the 0-1 loss is replaced by a convex surrogate loss for classification problems. In this paper, we investigate the validity of existing surrogate fairness constraints and propose a new surrogate fairness constraint called SLIDE, which is computationally feasible and asymptotically valid in the sense that the learned model satisfies the fairness constraint asymptotically and achieves a fast convergence rate. Numerical experiments confirm that the SLIDE works well for various benchmark datasets.