A Critical Review on the Use (and Misuse) of Differential Privacy in Machine Learning
ACM Computing Surveys · 31.08.2023 · 55 (8) · 1-16
DOI: 10.1145/3547139 ↗Kurz gefasst
Fachartikel in der Zeitschrift ACM Computing Surveys, online erschienen am 23.12.2022. Verfasst von Alberto Blanco-Justicia und 3 weiteren Autor:innen. Die Arbeit wurde bisher 89-mal zitiert (Crossref) und nennt 59 Literaturangaben. Gefördert von: European Commission, Norwegian Research Council, Government of Catalonia.
Abstract (Originalsprache)
We review the use of differential privacy (DP) for privacy protection in machine learning (ML). We show that, driven by the aim of preserving the accuracy of the learned models, DP-based ML implementations are so loose that they do not offer the ex ante privacy guarantees of DP. Instead, what they deliver is basically noise addition similar to the traditional (and often criticized) statistical disclosure control approach. Due to the lack of formal privacy guarantees, the actual level of privacy offered must be experimentally assessed ex post , which is done very seldom. In this respect, we present empirical results showing that standard anti-overfitting techniques in ML can achieve a better utility/privacy/efficiency tradeoff than DP.
Förderung
- European Commission
- Norwegian Research Council
- Government of Catalonia
- UK Research and Innovation
- UE
Zitieren
APA
Blanco-Justicia, A., Sánchez, D., Domingo-Ferrer, J., & Muralidhar, K. (2023). A Critical Review on the Use (and Misuse) of Differential Privacy in Machine Learning. ACM Computing Surveys, 55(8), 1-16. https://doi.org/10.1145/3547139
Vancouver
Blanco-Justicia A, Sánchez D, Domingo-Ferrer J, Muralidhar K. A Critical Review on the Use (and Misuse) of Differential Privacy in Machine Learning. ACM Computing Surveys. 2023;55(8):1-16. doi:10.1145/3547139
BibTeX
@article{BlancoJusticia2023,
author = {Blanco-Justicia, Alberto and Sánchez, David and Domingo-Ferrer, Josep and Muralidhar, Krishnamurty},
title = {A Critical Review on the Use (and Misuse) of Differential Privacy in Machine Learning},
journal = {ACM Computing Surveys},
year = {2023},
volume = {55},
number = {8},
pages = {1--16},
doi = {10.1145/3547139}
}
EUVIMED