Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead

被引:0
|
作者
Ghazi, Badih [1 ]
Kumar, Ravi [1 ]
Manurangsi, Pasin [1 ]
Pagh, Rasmus [1 ,2 ]
机构
[1] Google Res, Mountain View, CA 94043 USA
[2] It Univ Copenhagen, Copenhagen, Denmark
关键词
D O I
暂无
中图分类号
TP18 [人工智能理论];
学科分类号
081104 ; 0812 ; 0835 ; 1405 ;
摘要
Differential privacy (DP) is a formal notion for quantifying the privacy loss of algorithms. Algorithms in the central model of DP achieve high accuracy but make the strongest trust assumptions whereas those in the local DP model make the weakest trust assumptions but incur substantial accuracy loss. The shuffled DP model (Bittau et al., 2017; Erlingsson et al., 2019; Cheu et al., 2019) has recently emerged as a feasible middle ground between the central and local models, providing stronger trust assumptions than the former while promising higher accuracies than the latter. In this paper, we obtain practical communication-efficient algorithms in the shuffled DP model for two basic aggregation primitives used in machine learning: 1) binary summation, and 2) histograms over a moderate number of buckets. Our algorithms achieve accuracy that is arbitrarily close to that of central DP algorithms with an expected communication per user essentially matching what is needed without any privacy constraints! We demonstrate the practicality of our algorithms by experimentally comparing their performance to several widely-used protocols such as Randomized Response (Warner, 1965) and RAPPOR (Erlingsson et al., 2014).
引用
收藏
页数:10
相关论文
共 50 条
  • [1] Private Counting from Anonymous Messages: Near-Optimal Accuracy with Vanishing Communication Overhead
    Ghazi, Badih
    Kumar, Ravi
    Manurangsi, Pasin
    Pagh, Rasmus
    [J]. 25TH AMERICAS CONFERENCE ON INFORMATION SYSTEMS (AMCIS 2019), 2019,
  • [2] Near-Optimal Dispersion on Arbitrary Anonymous Graphs
    Kshemkalyani, Ajay D.
    Sharma, Gokarna
    [J]. Leibniz International Proceedings in Informatics, LIPIcs, 2022, 217
  • [3] Private Aggregation from Fewer Anonymous Messages
    Ghazi, Badih
    Manurangsi, Pasin
    Pagh, Rasmus
    Velingker, Ameya
    [J]. ADVANCES IN CRYPTOLOGY - EUROCRYPT 2020, PT II, 2020, 12106 : 798 - 827
  • [4] Near-Optimal Private Information Retrieval with Preprocessing
    Lazzaretti, Arthur
    Papamanthou, Charalampos
    [J]. THEORY OF CRYPTOGRAPHY, TCC 2023, PT II, 2023, 14370 : 406 - 435
  • [5] Near-Optimal Private and Scalable k-Clustering
    Cohen-Addad, Vincent
    Epasto, Alessandro
    Mirrokni, Vahab
    Narayanan, Shyam
    Zhong, Peilin
    [J]. ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35, NEURIPS 2022, 2022,
  • [6] Near-Optimal Self-stabilising Counting and Firing Squads
    Lenzen, Christoph
    Rybicki, Joel
    [J]. STABILIZATION, SAFETY, AND SECURITY OF DISTRIBUTED SYSTEMS, SSS 2016, 2016, 10083 : 263 - 280
  • [7] Near-optimal self-stabilising counting and firing squads
    Lenzen, Christoph
    Rybicki, Joel
    [J]. DISTRIBUTED COMPUTING, 2019, 32 (04) : 339 - 360
  • [8] Near-optimal self-stabilising counting and firing squads
    Christoph Lenzen
    Joel Rybicki
    [J]. Distributed Computing, 2019, 32 : 339 - 360
  • [9] Maintaining the Union of Unit Discs under Insertions with Near-Optimal Overhead
    Agarwal, Pankaj K.
    Cohen, Ravid
    Halperin, Dan
    Mulzer, Wolfgang
    [J]. ACM TRANSACTIONS ON ALGORITHMS, 2022, 18 (03)
  • [10] A near-optimal algorithm for differentially-private principal components
    Chaudhuri, Kamalika
    Sarwate, Anand D.
    Sinha, Kaushik
    [J]. Journal of Machine Learning Research, 2013, 14 : 2905 - 2943