Salvato in:
Dettagli Bibliografici
Autori principali: de Oliveira, Anderson Santana, Kaplan, Caelin, Mallat, Khawla, Chakraborty, Tanmay
Natura: Preprint
Pubblicazione: 2023
Soggetti:
Accesso online:https://arxiv.org/abs/2302.02910
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866914999758225408
author de Oliveira, Anderson Santana
Kaplan, Caelin
Mallat, Khawla
Chakraborty, Tanmay
author_facet de Oliveira, Anderson Santana
Kaplan, Caelin
Mallat, Khawla
Chakraborty, Tanmay
contents Recent works have shown that selecting an optimal model architecture suited to the differential privacy setting is necessary to achieve the best possible utility for a given privacy budget using differentially private stochastic gradient descent (DP-SGD)(Tramer and Boneh 2020; Cheng et al. 2022). In light of these findings, we empirically analyse how different fairness notions, belonging to distinct classes of statistical fairness criteria (independence, separation and sufficiency), are impacted when one selects a model architecture suitable for DP-SGD, optimized for utility. Using standard datasets from ML fairness literature, we show using a rigorous experimental protocol, that by selecting the optimal model architecture for DP-SGD, the differences across groups concerning the relevant fairness metrics (demographic parity, equalized odds and predictive parity) more often decrease or are negligibly impacted, compared to the non-private baseline, for which optimal model architecture has also been selected to maximize utility. These findings challenge the understanding that differential privacy will necessarily exacerbate unfairness in deep learning models trained on biased datasets.
format Preprint
id arxiv_https___arxiv_org_abs_2302_02910
institution arXiv
publishDate 2023
record_format arxiv
spellingShingle An Empirical Analysis of Fairness Notions under Differential Privacy
de Oliveira, Anderson Santana
Kaplan, Caelin
Mallat, Khawla
Chakraborty, Tanmay
Machine Learning
Recent works have shown that selecting an optimal model architecture suited to the differential privacy setting is necessary to achieve the best possible utility for a given privacy budget using differentially private stochastic gradient descent (DP-SGD)(Tramer and Boneh 2020; Cheng et al. 2022). In light of these findings, we empirically analyse how different fairness notions, belonging to distinct classes of statistical fairness criteria (independence, separation and sufficiency), are impacted when one selects a model architecture suitable for DP-SGD, optimized for utility. Using standard datasets from ML fairness literature, we show using a rigorous experimental protocol, that by selecting the optimal model architecture for DP-SGD, the differences across groups concerning the relevant fairness metrics (demographic parity, equalized odds and predictive parity) more often decrease or are negligibly impacted, compared to the non-private baseline, for which optimal model architecture has also been selected to maximize utility. These findings challenge the understanding that differential privacy will necessarily exacerbate unfairness in deep learning models trained on biased datasets.
title An Empirical Analysis of Fairness Notions under Differential Privacy
topic Machine Learning
url https://arxiv.org/abs/2302.02910