Balancing detectability and performance of attacks on the control channel of Markov Decision Processes

被引:0
|
作者
Russo, Alessio [1 ]
Proutiere, Alexandre [1 ]
机构
[1] KTH Royal Inst Technol, EECS Sch, Div Decis & Control Syst, Stockholm, Sweden
关键词
D O I
暂无
中图分类号
TP [自动化技术、计算机技术];
学科分类号
0812 ;
摘要
We investigate the problem of designing optimal stealthy poisoning attacks on the control channel of Markov decision processes (MDPs). This research is motivated by the recent interest of the research community for adversarial and poisoning attacks applied to MDPs, and reinforcement learning (RL) methods. The policies resulting from these methods have been shown to be vulnerable to attacks perturbing the observations of the decision-maker. In such an attack, drawing inspiration from adversarial examples used in supervised learning, the amplitude of the adversarial perturbation is limited according to some norm, with the hope that this constraint will make the attack imperceptible. However, such constraints do not grant any level of undetectability and do not take into account the dynamic nature of the underlying Markov process. In this paper, we propose a new attack formulation, based on information-theoretical quantities, that considers the objective of minimizing the detectability of the attack as well as the performance of the controlled process. We analyze the trade-off between the efficiency of the attack and its detectability. We conclude with examples and numerical simulations illustrating this trade-off.
引用
收藏
页码:2843 / 2850
页数:8
相关论文
共 50 条
  • [41] MARKOV DECISION-PROCESSES
    WHITE, CC
    WHITE, DJ
    EUROPEAN JOURNAL OF OPERATIONAL RESEARCH, 1989, 39 (01) : 1 - 16
  • [42] Algebraic Markov Decision Processes
    Perny, Patrice
    Spanjaard, Olivier
    Weng, Paul
    19TH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE (IJCAI-05), 2005, : 1372 - 1377
  • [43] Feature Markov Decision Processes
    Hutter, Marcus
    ARTIFICIAL GENERAL INTELLIGENCE PROCEEDINGS, 2009, 8 : 61 - 66
  • [44] Characterizing Markov decision processes
    Ratitch, B
    Precup, D
    MACHINE LEARNING: ECML 2002, 2002, 2430 : 391 - 404
  • [45] Absorbing Markov decision processes
    Dufour, Francois
    Prieto-Rumeau, Tomas
    ESAIM-CONTROL OPTIMISATION AND CALCULUS OF VARIATIONS, 2024, 30
  • [46] Logistic Markov Decision Processes
    Mladenov, Martin
    Boutilier, Craig
    Schuurmans, Dale
    Meshi, Ofer
    Elidan, Gal
    Lu, Tyler
    PROCEEDINGS OF THE TWENTY-SIXTH INTERNATIONAL JOINT CONFERENCE ON ARTIFICIAL INTELLIGENCE, 2017, : 2486 - 2493
  • [47] Quantile Markov Decision Processes
    Li, Xiaocheng
    Zhong, Huaiyang
    Brandeau, Margaret L.
    OPERATIONS RESEARCH, 2021, 70 (03) : 1428 - 1447
  • [48] Parallel markov decision processes
    Sucar, L. Enrique
    ADVANCES IN PROBABILISTIC GRAPHICAL MODELS, 2007, 213 : 295 - 309
  • [49] Configurable Markov Decision Processes
    Metelli, Alberto Maria
    Mutti, Mirco
    Restelli, Marcello
    INTERNATIONAL CONFERENCE ON MACHINE LEARNING, VOL 80, 2018, 80
  • [50] Possibilistic Markov decision processes
    Sabbadin, R
    ENGINEERING APPLICATIONS OF ARTIFICIAL INTELLIGENCE, 2001, 14 (03) : 287 - 300