Optimal Transport under Group Fairness Constraints

EurIPS 2025 Workshop UPLB Submission2 Authors

Keywords: optimal transport, fairness, matching

TL;DR: We devise fairness criteria for optimal transport between subgroups and provide efficient algorithms to obtain fair OT plans.

Track: Regular Paper

Abstract: Ensuring fairness in matching algorithms is a key challenge in allocating scarce resources and positions. Focusing on Optimal Transport (OT), we introduce a novel notion of group fairness requiring that the probability of matching two individuals from any two given groups in the OT plan matches a user-specified target. We develop two relaxation strategies of this constrained problem. The first one involves solving a penalized OT problem, for which we derive novel finite-sample complexity guarantees. Our second strategy leverages bi-level optimization to learn a ground cost that produces a fair OT solution, which can be reused to match new samples.

Submission Number: 2