"You Can’t Fix What You Can’t Measure": Privately Measuring Demographic Performance Disparities in Federated Learning

Marc Juarez Miro, Aleksandra Korolova

Research output: Chapter in Book/Report/Conference proceedingConference contribution

Abstract / Description of output

As in traditional machine learning models, models trained with federated learning may exhibit disparate performance across demographic groups. Model holders must identify these disparities to mitigate undue harm to the groups. However, measuring a model’s performance in a group requires access to information about group membership which, for privacy reasons, often has limited availability. We propose novel locally differentially private mechanisms to measure differences in performance across groups while protecting the privacy of group membership. To analyze the effectiveness of the mechanisms, we bound their error in estimating a disparity when optimized for a given privacy budget. Our results show that the error rapidly decreases for realistic numbers of participating clients, demonstrating that, contrary to what prior work suggested, protecting privacy is not necessarily in conflict with identifying performance disparities of federated models.
Original languageEnglish
Title of host publicationProceedings of Algorithmic Fairness through the Lens of Causality and Privacy: A hybrid NeurIPS 2022 Workshop
PublisherPMLR
Pages67-85
Number of pages19
Volume214
Publication statusPublished - 3 Dec 2023
EventAlgorithmic Fairness through the Lens of Causality and Privacy: A hybrid NeurIPS 2022 Workshop - New Orleans, United States
Duration: 3 Dec 20223 Dec 2022
https://www.afciworkshop.org/afcp2022

Publication series

NameProceedings of Machine Learning Research
ISSN (Print)2640-3498

Workshop

WorkshopAlgorithmic Fairness through the Lens of Causality and Privacy
Abbreviated titleAFCP
Country/TerritoryUnited States
CityNew Orleans
Period3/12/223/12/22
Internet address

Keywords / Materials (for Non-textual outputs)

  • differential privacy
  • algorithmic fairness
  • federated learning

Fingerprint

Dive into the research topics of '"You Can’t Fix What You Can’t Measure": Privately Measuring Demographic Performance Disparities in Federated Learning'. Together they form a unique fingerprint.

Cite this