Not All Learnable Distribution Classes are Privately Learnable

被引：0
|
作者：
Bun, Mark [1 ]
Kamath, Gautam [2 ,3 ]
Mouzakis, Argyris [2 ]
Singhal, Vikrant [4 ]
机构：
[1] Boston Univ, Boston, MA 02215 USA
[2] Univ Waterloo, Waterloo, ON, Canada
[3] Vector Inst, Toronto, ON, Canada
[4] OpenDP, Boston, MA USA
来源：
INTERNATIONAL CONFERENCE ON ALGORITHMIC LEARNING THEORY, VOL 237 | 2024年 / 237卷
基金：
加拿大自然科学与工程研究理事会;
关键词：
Differential privacy; distribution learning; lower bounds; SAMPLE COMPLEXITY; BOUNDS;
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：
We give an example of a class of distributions that is learnable in total variation distance with a finite number of samples, but not learnable under (epsilon,delta)-differential privacy. This refutes a conjecture of Ashtiani.
引用
收藏
页数：12
相关论文
共 50 条
  • [31] Learnable MFCCs for Speaker Verification
    Liu, Xuechen
    Sahidullah, Md
    Kinnunen, Tomi
    2021 IEEE INTERNATIONAL SYMPOSIUM ON CIRCUITS AND SYSTEMS (ISCAS), 2021,
  • [32] Are rational explosive solutions learnable?
    Kuang, Pei
    Yao, Yao
    ECONOMICS LETTERS, 2017, 157 : 62 - 66
  • [33] On the relative sizes of learnable sets
    Fortnow, L
    Freivalds, R
    Gasarch, WI
    Kummer, M
    Kurtz, SA
    Smith, CH
    Stephan, F
    THEORETICAL COMPUTER SCIENCE, 1998, 197 (1-2) : 139 - 156
  • [34] Learnable Embeddings of Program Spaces
    Krawiec, Krzysztof
    GENETIC PROGRAMMING, 2011, 6621 : 166 - 177
  • [35] Transformers with Learnable Activation Functions
    Fang, Haishuo
    Lee, Ji-Ung
    Moosavi, Nafise Sadat
    Gurevych, Iryna
    17TH CONFERENCE OF THE EUROPEAN CHAPTER OF THE ASSOCIATION FOR COMPUTATIONAL LINGUISTICS, EACL 2023, 2023, : 2382 - 2398
  • [36] APPLICATION PERIOD AS A LEARNABLE FACTOR
    HORTER, M
    ZEITSCHRIFT FUR TIERPSYCHOLOGIE-JOURNAL OF COMPARATIVE ETHOLOGY, 1977, 45 (03): : 256 - 287
  • [37] Regional and online learnable fields
    Schatten, R
    Goerke, NR
    Eckmiller, R
    2004 IEEE INTERNATIONAL JOINT CONFERENCE ON NEURAL NETWORKS, VOLS 1-4, PROCEEDINGS, 2004, : 793 - 798
  • [38] Learnable Programming: Blocks and Beyond
    Bau, David
    Gray, Jeff
    Kelleher, Caitlin
    Sheldon, Josh
    Turbak, Franklyn
    COMMUNICATIONS OF THE ACM, 2017, 60 (06) : 72 - 80
  • [39] Learnable Descriptors for Visual Search
    Migliorati, Andrea
    Fiandrotti, Attilio
    Francini, Gianluca
    Leonardi, Riccardo
    IEEE TRANSACTIONS ON IMAGE PROCESSING, 2021, 30 (30) : 80 - 91
  • [40] Semantically Adversarial Learnable Filters
    Shahin Shamsabadi, Ali
    Oh, Changjae
    Cavallaro, Andrea
    Shahin Shamsabadi, Ali (a.shahinshamsabadi@qmul.ac.uk), 1600, Institute of Electrical and Electronics Engineers Inc. (30): : 8075 - 8087
12345