Regression under demographic parity constraints via unlabeled post-processing

TL;DR

This paper introduces a theory-driven post-processing method for regression that ensures demographic parity without needing sensitive attributes during inference, applicable to unlabeled data and multi-class tasks.

Contribution

It presents a novel, theoretically grounded post-processing algorithm that guarantees demographic parity in regression without access to sensitive attributes at inference time.

Findings

Algorithm achieves demographic parity constraints effectively.

Finite-sample analysis validates the method's theoretical guarantees.

Experimental results confirm practical applicability and accuracy.

Abstract

We address the problem of performing regression while ensuring demographic parity, even without access to sensitive attributes during inference. We present a general-purpose post-processing algorithm that, using accurate estimates of the regression function and a sensitive attribute predictor, generates predictions that meet the demographic parity constraint. Our method involves discretization and stochastic minimization of a smooth convex function. It is suitable for online post-processing and multi-class classification tasks only involving unlabeled data for the post-processing. Unlike prior methods, our approach is fully theory-driven. We require precise control over the gradient norm of the convex function, and thus, we rely on more advanced techniques than standard stochastic gradient descent. Our algorithm is backed by finite-sample analysis and post-processing bounds, with experimental results validating our theoretical findings.