Robust Fair Clustering with Group Membership Uncertainty Sets

TL;DR

This paper introduces a robust fair clustering algorithm that handles noisy group membership data, providing theoretical guarantees and demonstrating superior empirical performance over existing methods.

Contribution

It proposes a new noise model for group memberships and an algorithm with provable robustness guarantees in fair clustering.

Findings

Algorithm achieves robustness against noisy group data.

The method outperforms existing baselines on real datasets.

Trade-off between robustness and clustering quality is demonstrated.

Abstract

We study the canonical fair clustering problem where each cluster is constrained to have close to population-level representation of each group. Despite significant attention, the salient issue of having incomplete knowledge about the group membership of each point has been superficially addressed. In this paper, we consider a setting where the assigned group memberships are noisy. We introduce a simple noise model that requires a small number of parameters to be given by the decision maker. We then present an algorithm for fair clustering with provable \emph{robustness} guarantees. Our framework enables the decision maker to trade off between the robustness and the clustering quality. Unlike previous work, our algorithms are backed by worst-case theoretical guarantees. Finally, we empirically verify the performance of our algorithm on real world datasets and show its superior performance over existing baselines.