Auditing Fairness by Betting

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Chugg, Ben, Cortes-Gomez, Santiago, Wilder, Bryan, Ramdas, Aaditya
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866908365728251904
author Chugg, Ben
Cortes-Gomez, Santiago
Wilder, Bryan
Ramdas, Aaditya
author_facet Chugg, Ben
Cortes-Gomez, Santiago
Wilder, Bryan
Ramdas, Aaditya
contents We provide practical, efficient, and nonparametric methods for auditing the fairness of deployed classification and regression models. Whereas previous work relies on a fixed-sample size, our methods are sequential and allow for the continuous monitoring of incoming data, making them highly amenable to tracking the fairness of real-world systems. We also allow the data to be collected by a probabilistic policy as opposed to sampled uniformly from the population. This enables auditing to be conducted on data gathered for another purpose. Moreover, this policy may change over time and different policies may be used on different subpopulations. Finally, our methods can handle distribution shift resulting from either changes to the model or changes in the underlying population. Our approach is based on recent progress in anytime-valid inference and game-theoretic statistics-the "testing by betting" framework in particular. These connections ensure that our methods are interpretable, fast, and easy to implement. We demonstrate the efficacy of our approach on three benchmark fairness datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2305_17570
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle Auditing Fairness by Betting
Chugg, Ben
Cortes-Gomez, Santiago
Wilder, Bryan
Ramdas, Aaditya
Machine Learning
Artificial Intelligence
Computers and Society
Applications
Methodology
We provide practical, efficient, and nonparametric methods for auditing the fairness of deployed classification and regression models. Whereas previous work relies on a fixed-sample size, our methods are sequential and allow for the continuous monitoring of incoming data, making them highly amenable to tracking the fairness of real-world systems. We also allow the data to be collected by a probabilistic policy as opposed to sampled uniformly from the population. This enables auditing to be conducted on data gathered for another purpose. Moreover, this policy may change over time and different policies may be used on different subpopulations. Finally, our methods can handle distribution shift resulting from either changes to the model or changes in the underlying population. Our approach is based on recent progress in anytime-valid inference and game-theoretic statistics-the "testing by betting" framework in particular. These connections ensure that our methods are interpretable, fast, and easy to implement. We demonstrate the efficacy of our approach on three benchmark fairness datasets.
title Auditing Fairness by Betting
topic Machine Learning
Artificial Intelligence
Computers and Society
Applications
Methodology
url https://arxiv.org/abs/2305.17570