2024/10/29 by Saeed Mahloujifar, Luca Melis, Mahloujifar, Saeed +3 · 1 voice · 12 citations
Computer Science · Social Sciences · #Accounting #Audit #Business #Computer science #Computer security #Cryptography and Data Security #Cryptography and Security (cs.CR) #Data mining #Differential privacy #FOS: Computer and information sciences #Internet privacy #Machine Learning (cs.LG) #Privacy, Security, and Data Protection #Privacy-Preserving Technologies in Data #cs.CR #cs.LG
paper · pdf · doi:10.48550/arxiv.2410.22235
published in arXiv (Cornell University) (Cornell University)
openalex publication_date 2024/10/29 · arxiv published 2024/10/29 · arxiv updated 2024/10/29 · openalex created_date 2025/10/10 · openalex updated_date 2026/07/28
Empirical auditing has emerged as a means of catching some of the flaws in the implementation of privacy-preserving algorithms. Existing auditing mechanisms, however, are either computationally inefficient requiring multiple runs of the machine learning algorithms or suboptimal in calculating an empirical privacy. In this work, we present a tight and efficient auditing procedure and analysis that can effectively assess the privacy of mechanisms. Our approach is efficient; similar to the recent work of Steinke, Nasr, and Jagielski (2023), our auditing procedure leverages the randomness of examples in the input dataset and requires only a single run of the target mechanism. And it is more accurate; we provide a novel analysis that enables us to achieve tight empirical privacy estimates by using the hypothesized f-DP curve of the mechanism, which provides a more accurate measure of privacy than the traditional ε,δ differential privacy parameters. We use our auditing procure and analysis to obtain empirical privacy, demonstrating that our auditing procedure delivers tighter privacy estimates.