The Complexity of Computing the Optimal Composition of Differential Privacy

被引:41
|
作者
Murtagh, Jack [1 ]
Vadhan, Salil [1 ]
机构
[1] Harvard Univ, John A Paulson Sch Engn & Appl Sci, Ctr Res Computat & Soc, Cambridge, MA 02138 USA
来源
THEORY OF CRYPTOGRAPHY, TCC 2016-A, PT I | 2016年 / 9562卷
关键词
Differential privacy; Composition; Computational complexity; Approximation algorithms; NOISE;
D O I
10.1007/978-3-662-49096-9_7
中图分类号
TP301 [理论、方法];
学科分类号
081202 ;
摘要
In the study of differential privacy, composition theorems (starting with the original paper of Dwork, McSherry, Nissim, and Smith (TCC'06)) bound the degradation of privacy when composing several differentially private algorithms. Kairouz, Oh, and Viswanath (ICML'15) showed how to compute the optimal bound for composing k arbitrary (epsilon, delta)-differentially private algorithms. We characterize the optimal composition for the more general case of k arbitrary (epsilon(1), delta(1)), ... , (epsilon(k), delta(k))-differentially private algorithms where the privacy parameters may differ for each algorithm in the composition. We show that computing the optimal composition in general is #P-complete. Since computing optimal composition exactly is infeasible (unless FP=#P), we give an approximation algorithm that computes the composition to arbitrary accuracy in polynomial time. The algorithm is a modification of Dyer's dynamic programming approach to approximately counting solutions to knapsack problems (STOC'03).
引用
收藏
页码:157 / 175
页数:19
相关论文
共 50 条
  • [41] Discrete Offset-Symmetric Gaussians for Differential Privacy
    Korki, Mehdi
    Sadeghi, Parastoo
    IEEE SIGNAL PROCESSING LETTERS, 2024, 31 : 1915 - 1919
  • [42] Privacy-Preserving Federated Learning for Industrial Edge Computing via Hybrid Differential Privacy and Adaptive Compression
    Jiang, Bin
    Li, Jianqiang
    Wang, Huihui
    Song, Houbing
    IEEE TRANSACTIONS ON INDUSTRIAL INFORMATICS, 2023, 19 (02) : 1136 - 1144
  • [43] Optimal Privacy Preserving in Wireless Federated Learning over Mobile Edge Computing
    Nguyen, Hai M.
    Chu, Nam H.
    Nguyen, Diep N.
    Dinh Thai Hoang
    Minh Hoang Ha
    Dutkiewicz, Eryk
    ICC 2023-IEEE INTERNATIONAL CONFERENCE ON COMMUNICATIONS, 2023, : 2000 - 2006
  • [44] Helmholtz machine with differential privacy
    Hu, Junying
    Sun, Kai
    Zhang, Hai
    INFORMATION SCIENCES, 2022, 613 : 888 - 903
  • [45] Data Loss and Reconstruction of Location Differential Privacy Protection Based on Edge Computing
    Jing, Weipeng
    Miao, Qiucheng
    Song, Houbing
    Chen, Xuebin
    IEEE ACCESS, 2019, 7 : 75890 - 75900
  • [46] A Local Differential Privacy Hybrid Data Clustering Iterative Algorithm for Edge Computing
    Zhou, Yousheng
    Wang, Zhonghan
    Liu, Yuanni
    CHINESE JOURNAL OF ELECTRONICS, 2024, 33 (06) : 1421 - 1434
  • [47] Differential Privacy Preserving of Training Model in Wireless Big Data with Edge Computing
    Du, Miao
    Wang, Kun
    Xia, Zhuoqun
    Zhang, Yan
    IEEE TRANSACTIONS ON BIG DATA, 2020, 6 (02) : 283 - 295
  • [48] Releasing Correlated Trajectories: Towards High Utility and Optimal Differential Privacy
    Ou, Lu
    Qin, Zheng
    Liao, Shaolin
    Hong, Yuan
    Jia, Xiaohua
    IEEE TRANSACTIONS ON DEPENDABLE AND SECURE COMPUTING, 2020, 17 (05) : 1109 - 1123
  • [49] Lower Bounds in Differential Privacy
    De, Anindya
    THEORY OF CRYPTOGRAPHY (TCC 2012), 2012, 7194 : 321 - 338
  • [50] A Differential Privacy Collaborative Deep Learning Algorithm in Pervasive Edge Computing Environment
    Zhang, Dayin
    Chen, Xiaojun
    Shi, Jinqiao
    Wang, Dakui
    Zeng, Shuai
    2021 IEEE 20TH INTERNATIONAL CONFERENCE ON TRUST, SECURITY AND PRIVACY IN COMPUTING AND COMMUNICATIONS (TRUSTCOM 2021), 2021, : 347 - 354