Optimizing Error of High-Dimensional Statistical Queries Under Differential Privacy
Journal of Privacy and Confidentiality · 31.08.2023 · 13 (1)
DOI: 10.29012/jpc.791 ↗Kurz gefasst
Fachartikel in der Zeitschrift Journal of Privacy and Confidentiality, online erschienen am 31.08.2023. Verfasst von Ryan McKenna und 3 weiteren Autor:innen. Die Arbeit wurde bisher 7-mal zitiert (Crossref). Der Volltext ist frei zugänglich (Open Access, CC BY-NC-ND 4.0). Gefördert von: Division of Computer and Network Systems, Office of Advanced Cyberinfrastructure, Defense Advanced Research Projects Agency.
Abstract (Originalsprache)
In this work we describe the High-Dimensional Matrix Mechanism (HDMM),a differentially private algorithm for answering a workload of predicate counting queries. HDMM represents query workloads using a compact implicit matrix representation and exploits this representation to efficiently optimize over (a subset of) the space of differentially private algorithms for one that is unbiased and answers the input query workload with low expected error. HDMM can be deployed for both ϵ-differential privacy (with Laplace noise) and (ϵ, δ)-differential privacy (with Gaussian noise), although the core techniques are slightly different for each. We demonstrate empirically that HDMM can efficiently answer queries with lower expected error than state-of-the-art techniques, and in some cases, it nearly matches existing lower bounds for the particular class of mechanisms we consider.
Förderung
- Division of Computer and Network Systems
- Office of Advanced Cyberinfrastructure
- Defense Advanced Research Projects Agency
- Space and Naval Warfare Systems Command
Zitieren
APA
McKenna, R., Miklau, G., Hay, M., & Machanavajjhala, A. (2023). Optimizing Error of High-Dimensional Statistical Queries Under Differential Privacy. Journal of Privacy and Confidentiality, 13(1). https://doi.org/10.29012/jpc.791
Vancouver
McKenna R, Miklau G, Hay M, Machanavajjhala A. Optimizing Error of High-Dimensional Statistical Queries Under Differential Privacy. Journal of Privacy and Confidentiality. 2023;13(1). doi:10.29012/jpc.791
BibTeX
@article{McKenna2023,
author = {McKenna, Ryan and Miklau, Gerome and Hay, Michael and Machanavajjhala, Ashwin},
title = {Optimizing Error of High-Dimensional Statistical Queries Under Differential Privacy},
journal = {Journal of Privacy and Confidentiality},
year = {2023},
volume = {13},
number = {1},
doi = {10.29012/jpc.791}
}
EUVIMED