Cargando…

A Maximal Correlation Framework for Fair Machine Learning

As machine learning algorithms grow in popularity and diversify to many industries, ethical and legal concerns regarding their fairness have become increasingly relevant. We explore the problem of algorithmic fairness, taking an information–theoretic view. The maximal correlation framework is introd...

Descripción completa

Detalles Bibliográficos
Autores principales: Lee, Joshua, Bu, Yuheng, Sattigeri, Prasanna, Panda, Rameswar, Wornell, Gregory W., Karlinsky, Leonid, Schmidt Feris, Rogerio
Formato: Online Artículo Texto
Lenguaje:English
Publicado: MDPI 2022
Materias:
Acceso en línea:https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9027582/
https://www.ncbi.nlm.nih.gov/pubmed/35455124
http://dx.doi.org/10.3390/e24040461
_version_ 1784691402018390016
author Lee, Joshua
Bu, Yuheng
Sattigeri, Prasanna
Panda, Rameswar
Wornell, Gregory W.
Karlinsky, Leonid
Schmidt Feris, Rogerio
author_facet Lee, Joshua
Bu, Yuheng
Sattigeri, Prasanna
Panda, Rameswar
Wornell, Gregory W.
Karlinsky, Leonid
Schmidt Feris, Rogerio
author_sort Lee, Joshua
collection PubMed
description As machine learning algorithms grow in popularity and diversify to many industries, ethical and legal concerns regarding their fairness have become increasingly relevant. We explore the problem of algorithmic fairness, taking an information–theoretic view. The maximal correlation framework is introduced for expressing fairness constraints and is shown to be capable of being used to derive regularizers that enforce independence and separation-based fairness criteria, which admit optimization algorithms for both discrete and continuous variables that are more computationally efficient than existing algorithms. We show that these algorithms provide smooth performance–fairness tradeoff curves and perform competitively with state-of-the-art methods on both discrete datasets (COMPAS, Adult) and continuous datasets (Communities and Crimes).
format Online
Article
Text
id pubmed-9027582
institution National Center for Biotechnology Information
language English
publishDate 2022
publisher MDPI
record_format MEDLINE/PubMed
spelling pubmed-90275822022-04-23 A Maximal Correlation Framework for Fair Machine Learning Lee, Joshua Bu, Yuheng Sattigeri, Prasanna Panda, Rameswar Wornell, Gregory W. Karlinsky, Leonid Schmidt Feris, Rogerio Entropy (Basel) Article As machine learning algorithms grow in popularity and diversify to many industries, ethical and legal concerns regarding their fairness have become increasingly relevant. We explore the problem of algorithmic fairness, taking an information–theoretic view. The maximal correlation framework is introduced for expressing fairness constraints and is shown to be capable of being used to derive regularizers that enforce independence and separation-based fairness criteria, which admit optimization algorithms for both discrete and continuous variables that are more computationally efficient than existing algorithms. We show that these algorithms provide smooth performance–fairness tradeoff curves and perform competitively with state-of-the-art methods on both discrete datasets (COMPAS, Adult) and continuous datasets (Communities and Crimes). MDPI 2022-03-26 /pmc/articles/PMC9027582/ /pubmed/35455124 http://dx.doi.org/10.3390/e24040461 Text en © 2022 by the authors. https://creativecommons.org/licenses/by/4.0/Licensee MDPI, Basel, Switzerland. This article is an open access article distributed under the terms and conditions of the Creative Commons Attribution (CC BY) license (https://creativecommons.org/licenses/by/4.0/).
spellingShingle Article
Lee, Joshua
Bu, Yuheng
Sattigeri, Prasanna
Panda, Rameswar
Wornell, Gregory W.
Karlinsky, Leonid
Schmidt Feris, Rogerio
A Maximal Correlation Framework for Fair Machine Learning
title A Maximal Correlation Framework for Fair Machine Learning
title_full A Maximal Correlation Framework for Fair Machine Learning
title_fullStr A Maximal Correlation Framework for Fair Machine Learning
title_full_unstemmed A Maximal Correlation Framework for Fair Machine Learning
title_short A Maximal Correlation Framework for Fair Machine Learning
title_sort maximal correlation framework for fair machine learning
topic Article
url https://www.ncbi.nlm.nih.gov/pmc/articles/PMC9027582/
https://www.ncbi.nlm.nih.gov/pubmed/35455124
http://dx.doi.org/10.3390/e24040461
work_keys_str_mv AT leejoshua amaximalcorrelationframeworkforfairmachinelearning
AT buyuheng amaximalcorrelationframeworkforfairmachinelearning
AT sattigeriprasanna amaximalcorrelationframeworkforfairmachinelearning
AT pandarameswar amaximalcorrelationframeworkforfairmachinelearning
AT wornellgregoryw amaximalcorrelationframeworkforfairmachinelearning
AT karlinskyleonid amaximalcorrelationframeworkforfairmachinelearning
AT schmidtferisrogerio amaximalcorrelationframeworkforfairmachinelearning
AT leejoshua maximalcorrelationframeworkforfairmachinelearning
AT buyuheng maximalcorrelationframeworkforfairmachinelearning
AT sattigeriprasanna maximalcorrelationframeworkforfairmachinelearning
AT pandarameswar maximalcorrelationframeworkforfairmachinelearning
AT wornellgregoryw maximalcorrelationframeworkforfairmachinelearning
AT karlinskyleonid maximalcorrelationframeworkforfairmachinelearning
AT schmidtferisrogerio maximalcorrelationframeworkforfairmachinelearning