Fair Soft Clustering
Rune D. Kjærsgaard, Pekka Parviainen, Saket Saurabh, Madhumita Kundu, Line Clemmensen
Abstract
Scholars in the machine learning community have recently focused on analyzing the fairness of learning models, including clustering algorithms. In this work we study fair clustering in a probabilistic (soft) setting, where observations may belong to several clusters determined by probabilities. We introduce new probabilistic fairness metrics, which generalize and extend existing non-probabilistic fairness frameworks and propose an algorithm for obtaining a fair probabilistic cluster solution from a data representation known as a fairlet decomposition. Finally, we demonstrate our proposed fairness metrics and algorithm by constructing a fair Gaussian mixture model on three real-world datasets. We achieve this by identifying balanced micro-clusters which minimize the distances induced by the model, and on which traditional clustering can be performed while ensuring the fairness of the solution.
BibTeX
@InProceedings{pmlr-v238-kjaersgaard24a,
title = {Fair Soft Clustering},
author = {Kj{\ae}rsgaard, Rune D. and Parviainen, Pekka and Saurabh, Saket and Kundu, Madhumita and Clemmensen, Line},
booktitle = {Proceedings of The 27th International Conference on Artificial Intelligence and Statistics},
pages = {1270--1278},
year = {2024},
editor = {Dasgupta, Sanjoy and Mandt, Stephan and Li, Yingzhen},
volume = {238},
series = {Proceedings of Machine Learning Research},
month = {02--04 May},
publisher = {PMLR},
pdf = {https://proceedings.mlr.press/v238/kjaersgaard24a/kjaersgaard24a.pdf},
url = {https://proceedings.mlr.press/v238/kjaersgaard24a.html},
abstract = {Scholars in the machine learning community have recently focused on analyzing the fairness of learning models, including clustering algorithms. In this work we study fair clustering in a probabilistic (soft) setting, where observations may belong to several clusters determined by probabilities. We introduce new probabilistic fairness metrics, which generalize and extend existing non-probabilistic fairness frameworks and propose an algorithm for obtaining a fair probabilistic cluster solution from a data representation known as a fairlet decomposition. Finally, we demonstrate our proposed fairness metrics and algorithm by constructing a fair Gaussian mixture model on three real-world datasets. We achieve this by identifying balanced micro-clusters which minimize the distances induced by the model, and on which traditional clustering can be performed while ensuring the fairness of the solution.}
}