Skip to main content

Research Repository

Advanced Search

Measuring AI Fairness in a Continuum Maintaining Nuances: A Robustness Case Study

Paxton, Kuniko; Aslansefat, Koorosh; Thakker, Dhavalkumar; Papadopoulos, Yiannis

Authors

Kuniko Paxton



Abstract

As machine learning is increasingly making decisions about hiring or healthcare, we want AI to treat ethnic and socioeconomic groups fairly. Fairness is currently measured by comparing the average accuracy of reasoning across groups. We argue that improved measurement is possible on a continuum and without averaging, with the advantage that nuances could be observed within groups. Through the example of skin cancer diagnosis, we illustrate a new statistical method that works on multidimensional data and treats fairness in a continuum. We outline this new approach and focus on its robustness against three types of adversarial attacks. Indeed, such attacks can influence data in ways that may cause different levels of misdiagnosis for different skin tones, thereby distorting fairness. Our results reveal nuances that would not be evident in a strictly categorical approach.

Citation

Paxton, K., Aslansefat, K., Thakker, D., & Papadopoulos, Y. (2024). Measuring AI Fairness in a Continuum Maintaining Nuances: A Robustness Case Study. IEEE Internet Computing, 28(5), 11-19. https://doi.org/10.1109/MIC.2024.3450815

Journal Article Type Article
Acceptance Date Sep 1, 2024
Online Publication Date Aug 29, 2024
Publication Date Sep 1, 2024
Deposit Date Dec 10, 2024
Publicly Available Date Dec 12, 2024
Journal IEEE Internet Computing
Print ISSN 1089-7801
Electronic ISSN 1941-0131
Publisher Institute of Electrical and Electronics Engineers
Peer Reviewed Peer Reviewed
Volume 28
Issue 5
Pages 11-19
DOI https://doi.org/10.1109/MIC.2024.3450815
Keywords Robustness; Accuracy; Ethnicity; Machine learning; Artificial intelligence; Ethics; Social implications of technology; Sociotechnical systems; Cultural differences
Public URL https://hull-repository.worktribe.com/output/4834268

Files

Accepted manuscript (2 Mb)
PDF

Copyright Statement
Copyright © 2024 IEEE. Personal use of this material is permitted. Permission from IEEE must be obtained for all other uses, in any current or future media, including reprinting/republishing this material for advertising or promotional purposes, creating new collective works, for resale or redistribution to servers or lists, or reuse of any copyrighted component of this work in other works.




You might also like



Downloadable Citations