Fair Clustering with Clusterlets

TL;DR

This paper introduces clusterlet-based fuzzy clustering algorithms that enhance fairness in clustering by matching small, fair clusters, balancing fairness with cohesion and overlap through parameter tuning.

Contribution

It proposes a novel clusterlet-based fuzzy clustering method that simplifies achieving fair clustering by leveraging clusterlet matching and regularization.

Findings

High fairness achieved with simple matching strategies

Parameter tuning balances fairness, cohesion, and overlap

Empirical results validate effectiveness of the proposed algorithms

Abstract

Given their widespread usage in the real world, the fairness of clustering methods has become of major interest. Theoretical results on fair clustering show that fairness enjoys transitivity: given a set of small and fair clusters, a trivial centroid-based clustering algorithm yields a fair clustering. Unfortunately, discovering a suitable starting clustering can be computationally expensive, rather complex or arbitrary. In this paper, we propose a set of simple \emph{clusterlet}-based fuzzy clustering algorithms that match single-class clusters, optimizing fair clustering. Matching leverages clusterlet distance, optimizing for classic clustering objectives, while also regularizing for fairness. Empirical results show that simple matching strategies are able to achieve high fairness, and that appropriate parameter tuning allows to achieve high cohesion and low overlap.