Browse by author
Lookup NU author(s): Carlos Vladimiro Gonzalez Zelayn, Professor Paolo MissierORCiD
This work is licensed under a Creative Commons Attribution 4.0 International License (CC BY 4.0).
Copyright © 2023 held by the owner/author(s).Privacy protection for personal data and fairness in automated decisions are fundamental requirements for responsible Machine Learning. Both may be enforced through data preprocessing and share a common target: data should remain useful for a task, while becoming uninformative of the sensitive information. The intrinsic connection between privacy and fairness implies that modifications performed to guarantee one of these goals, may have an effect on the other, e.g., hiding a sensitive attribute from a classification algorithm might prevent a biased decision rule having such attribute as a criterion. This work resides at the intersection of algorithmic fairness and privacy. We show how the two goals are compatible, and may be simultaneously achieved, with a small loss in predictive performance. Our results are competitive with both state-of-the-art fairness correcting algorithms and hybrid privacy-fairness methods. Experiments were performed on three widely used benchmark datasets: Adult Income, COMPAS, and German Credit.
Author(s): Gonzalez-Zelaya V, Salas J, Megias D, Missier P
Publication type: Article
Publication status: Published
Journal: ACM Transactions on Knowledge Discovery from Data
Year: 2024
Volume: 18
Issue: 3
Print publication date: 01/04/2024
Online publication date: 09/12/2023
Acceptance date: 18/08/2023
Date deposited: 30/01/2024
ISSN (print): 1556-4681
ISSN (electronic): 1556-472X
Publisher: Association for Computing Machinery
URL: https://doi.org/10.1145/3617377
DOI: 10.1145/3617377
Data Access Statement: The datasets on which our experiments were run are available at: Adult Income: https://archive.ics.uci.edu/ml/datasets/adult COMPAS: https://github.com/propublica/compas-analysis German Credit: https://archive.ics.uci.edu/ml/datasets/statlog+(german+credit+data) They were all prepared using the data_cleanup.ipynb notebook available at the project’s repository.
Altmetrics provided by Altmetric
