From Pseudorandomness to Multi-Group Fairness and Back
Machine Learning
2025-07-10 v4 Computational Complexity
Abstract
We identify and explore connections between the recent literature on multi-group fairness for prediction algorithms and the pseudorandomness notions of leakage-resilience and graph regularity. We frame our investigation using new variants of multicalibration based on statistical distance and closely related to the concept of outcome indistinguishability. Adopting this perspective leads us not only to new, more efficient algorithms for multicalibration, but also to our graph theoretic results and a proof of a novel hardcore lemma for real-valued functions.
Cite
@article{arxiv.2301.08837,
title = {From Pseudorandomness to Multi-Group Fairness and Back},
author = {Cynthia Dwork and Daniel Lee and Huijia Lin and Pranay Tankala},
journal= {arXiv preprint arXiv:2301.08837},
year = {2025}
}