Robust Fair Clustering with Group Membership Uncertainty Sets

Fuente: arXiv
Enregistré dans:
Détails bibliographiques
Auteurs principaux: Duppala, Sharmila, Luque, Juan, Dickerson, John P., Esmaeili, Seyed A.
Format: Preprint
Publié: 2024
Sujets:
Accès en ligne:
Tags: Ajouter un tag
Pas de tags, Soyez le premier à ajouter un tag!
_version_ 1866915026905858048
author Duppala, Sharmila
Luque, Juan
Dickerson, John P.
Esmaeili, Seyed A.
author_facet Duppala, Sharmila
Luque, Juan
Dickerson, John P.
Esmaeili, Seyed A.
contents 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.
format Preprint
id arxiv_https___arxiv_org_abs_2406_00599
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Robust Fair Clustering with Group Membership Uncertainty Sets
Duppala, Sharmila
Luque, Juan
Dickerson, John P.
Esmaeili, Seyed A.
Machine Learning
Artificial Intelligence
Computers and Society
Data Structures and Algorithms
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.
title Robust Fair Clustering with Group Membership Uncertainty Sets
topic Machine Learning
Artificial Intelligence
Computers and Society
Data Structures and Algorithms
url https://arxiv.org/abs/2406.00599