FairICP: Encouraging Equalized Odds via Inverse Conditional Permutation

Fuente: arXiv
Saved in:
Bibliographic Details
Main Authors: Lai, Yuheng, Guan, Leying
Format: Preprint
Published: 2024
Subjects:
Online Access:
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1866912417612562432
author Lai, Yuheng
Guan, Leying
author_facet Lai, Yuheng
Guan, Leying
contents $\textit{Equalized odds}$, an important notion of algorithmic fairness, aims to ensure that sensitive variables, such as race and gender, do not unfairly influence the algorithm's prediction when conditioning on the true outcome. Despite rapid advancements, current research primarily focuses on equalized odds violations caused by a single sensitive attribute, leaving the challenge of simultaneously accounting for multiple attributes under-addressed. We bridge this gap by introducing an in-processing fairness-aware learning approach, FairICP, which integrates adversarial learning with a novel inverse conditional permutation scheme. FairICP offers a flexible and efficient scheme to promote equalized odds under fairness conditions described by complex and multi-dimensional sensitive attributes. The efficacy and adaptability of our method are demonstrated through both simulation studies and empirical analyses of real-world datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2404_05678
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle FairICP: Encouraging Equalized Odds via Inverse Conditional Permutation
Lai, Yuheng
Guan, Leying
Machine Learning
Computers and Society
$\textit{Equalized odds}$, an important notion of algorithmic fairness, aims to ensure that sensitive variables, such as race and gender, do not unfairly influence the algorithm's prediction when conditioning on the true outcome. Despite rapid advancements, current research primarily focuses on equalized odds violations caused by a single sensitive attribute, leaving the challenge of simultaneously accounting for multiple attributes under-addressed. We bridge this gap by introducing an in-processing fairness-aware learning approach, FairICP, which integrates adversarial learning with a novel inverse conditional permutation scheme. FairICP offers a flexible and efficient scheme to promote equalized odds under fairness conditions described by complex and multi-dimensional sensitive attributes. The efficacy and adaptability of our method are demonstrated through both simulation studies and empirical analyses of real-world datasets.
title FairICP: Encouraging Equalized Odds via Inverse Conditional Permutation
topic Machine Learning
Computers and Society
url https://arxiv.org/abs/2404.05678