Adversarial Risk via Optimal Transport and Optimal Couplings

被引:14
|
作者
Pydi, Muni Sreenivas [1 ]
Jog, Varun [2 ]
机构
[1] Univ Wisconsin, Dept Elect & Comp Engn, 1415 Johnson Dr, Madison, WI 53706 USA
[2] Univ Cambridge, Dept Pure Math & Math Stat, Cambridge CB3 0WB, England
关键词
Couplings; Standards; Measurement; Kernel; Perturbation methods; Loss measurement; Q measurement; Machine learning; statistical learning; robustness; couplings; information theory; DEEP NEURAL-NETWORKS; ROBUST; ALGORITHMS; GO;
D O I
10.1109/TIT.2021.3100107
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
Modern machine learning algorithms perform poorly on adversarially manipulated data. Adversarial risk quantifies the error of classifiers in adversarial settings; adversarial classifiers minimize adversarial risk. In this paper, we analyze adversarial risk and adversarial classifiers from an optimal transport perspective. We show that the optimal adversarial risk for binary classification with 0-1 loss is determined by an optimal transport cost between the probability distributions of the two classes. We develop optimal transport plans (probabilistic couplings) for univariate distributions such as the normal, the uniform, and the triangular distribution. We also derive optimal adversarial classifiers in these settings. Our analysis leads to algorithm-independent fundamental limits on adversarial risk, which we calculate for several real-world datasets. We extend our results to general loss functions under convexity and smoothness assumptions.
引用
收藏
页码:6031 / 6052
页数:22
相关论文
共 50 条
  • [21] Margin-aware Adversarial Domain Adaptation with Optimal Transport
    Dhouib, Sofien
    Redko, Ievgen
    Lartizien, Carole
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 119, 2020, 119
  • [22] Optimal Markovian Couplings and Applications
    Chen Mufa Department of Mathematics Beijing Normal University Beijing
    Acta Mathematica Sinica,English Series, 1994, (03) : 260 - 275
  • [23] Optimal Efficiency-Envy Trade-Off via Optimal Transport
    Yin, Steven
    Kroer, Christian
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 35 (NEURIPS 2022), 2022,
  • [24] HOT-GAN: Hilbert Optimal Transport for Generative Adversarial Network
    Li, Qian
    Wang, Zhichao
    Xia, Haiyang
    Li, Gang
    Cao, Yanan
    Yao, Lina
    Xu, Guandong
    IEEE TRANSACTIONS ON NEURAL NETWORKS AND LEARNING SYSTEMS, 2024, : 1 - 14
  • [25] Learning to Match via Inverse Optimal Transport
    Li, Ruilin
    Ye, Xiaojing
    Zhou, Haomin
    Zha, Hongyuan
    JOURNAL OF MACHINE LEARNING RESEARCH, 2019, 20
  • [26] Geometric Dataset Distances via Optimal Transport
    Alvarez-Melis, David
    Fusi, Nicolo
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 33, NEURIPS 2020, 2020, 33
  • [27] Learning to Count via Unbalanced Optimal Transport
    Ma, Zhiheng
    Wei, Xing
    Hong, Xiaopeng
    Lin, Hui
    Qiu, Yunfeng
    Gong, Yihong
    THIRTY-FIFTH AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE, THIRTY-THIRD CONFERENCE ON INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE AND THE ELEVENTH SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2021, 35 : 2319 - 2327
  • [28] Visual Prompting via Partial Optimal Transport
    Zheng, Mengyu
    Hao, Zhiwei
    Tang, Yehui
    Xu, Chang
    COMPUTER VISION-ECCV 2024, PT XXXV, 2025, 15093 : 1 - 18
  • [29] Oversampling for Imbalanced Data via Optimal Transport
    Yan, Yuguang
    Tan, Mingkui
    Xu, Yanwu
    Cao, Jiezhang
    Ng, Michael
    Min, Huaqing
    Wu, Qingyao
    THIRTY-THIRD AAAI CONFERENCE ON ARTIFICIAL INTELLIGENCE / THIRTY-FIRST INNOVATIVE APPLICATIONS OF ARTIFICIAL INTELLIGENCE CONFERENCE / NINTH AAAI SYMPOSIUM ON EDUCATIONAL ADVANCES IN ARTIFICIAL INTELLIGENCE, 2019, : 5605 - 5612
  • [30] Accelerating Motion Planning via Optimal Transport
    Le, An T.
    Chalvatzaki, Georgia
    Biess, Armin
    Peters, Jan
    ADVANCES IN NEURAL INFORMATION PROCESSING SYSTEMS 36, NEURIPS 2023, 2023,