Dynamic Cheap Talk for Robust Adversarial Learning
Authors: 
Zuxing Li, György DánAuthors Info & Claims
Pages 297 - 309
Abstract
Robust adversarial learning is considered in the context of closed-loop control with adversarial signaling in this paper. Due to the nature of incomplete information of the control agent about the environment, the belief-dependent signaling game formulation is introduced in the dynamic system and a dynamic cheap talk game is formulated with belief-dependent strategies for both players. We show that the dynamic cheap talk game can further be reformulated as a particular stochastic game, where the states are beliefs of the environment and the actions are the adversarial manipulation strategies and control strategies. Furthermore, the bisimulation metric is proposed and studied for the dynamic cheap talk game, which provides an upper bound on the difference between values of different initial beliefs in the zero-sum equilibrium.
References
[1]
Goodfellow I, Bengio Y, and Courville A Deep Learning 2016 Cambridge MIT Press
[2]
Pinto, L., Davidson, J., Sukthankar, R., Gupta, A.: Robust adversarial reinforcement learning. In: Proceedings of the 34th International Conference on Machine Learning, pp. 1–10 (2017)
[3]
Pan, X., Seita, D., Gao, Y., Canny, J.: Risk averse robust adversarial reinforcement learning. In: Proceedings of the IEEE International Conference on Robotics and Automation (2019)
[4]
Huang, S., Papernot, N., Goodfellow, I., Duan, Y., Abbeel, P.: Adversarial attacks on neural network policies. arXiv:1702.02284 (2016)
[5]
Szegedy, C., et al.: Intriguing properties of neural networks. arXiv:1312.6199v4 (2013)
[6]
Dong, Y., et al.: Boosting adversarial attacks with momentum. In: Proceedings of the 2018 Conference on Computer Vision and Pattern Recognition, pp. 9185–9193 (2018)
[7]
Papernot, N., McDaniel, P., Jha, S., Fredrikson, M., Celik, Z.B., Swami, A.: The limitations of deep learning in adversarial settings. In: Proceedings of IEEE European Symposium on Security and Privacy (2016)
[8]
Mnih V et al. Human-level control through deep reinforcement learning Nature 2015 518 7540 529-533
[9]
Shapley LS Stochastic games Proc. Natl. Acad. Sci. 1953 39 10 1095-1100
[10]
He X and Dai H Dynamic Games for Network Security 2018 Cham Springer
[11]
He X, Dai H, and Ning P Faster learning and adaptation in security games by exploiting information asymmetry IEEE Trans. Sig. Process. 2016 64 13 3429-3443
[12]
Busoniu L, Babuska R, and Schutter B A comprehensive survey of multiagent reinforcement learning IEEE Trans. Syst. Man Cybern. - Part C 2008 38 2 156-172
[13]
Horák, K., Bošanský, B., Pěchouček, M.: Heuristic search value iteration for one-sided partially observable stochastic games. In: Proceedings of the 31st AAAI Conference on Artificial Intelligence, pp. 558–564 (2017)
[14]
Ferns, N., Panangaden, P., Precup, D.: Metrics for finite Markov decision processes. In: Proceedings of the 20th Conference on Uncertainty in Artificial Intelligence, pp. 162–169 (2004)
[15]
Alfaro, L., Majumdar, R., Raman, V., Stoelinga, M.: Game relations and metrics. In: Proceedings of the 22nd Annual IEEE Symposium on Logic in Computer Science, pp. 99–108 (2007)
[16]
Chatterjee K, Alfaro L, Majumdar R, and Raman V Algorithms for game metrics Logic. Methods Comput. Sci. 2010 6 3:13 1-27
[17]
Crawford VP and Sobel J Strategic information transmission Econometrica 1982 50 6 1431-1451
Index Terms
- Dynamic Cheap Talk for Robust Adversarial Learning
Index terms have been assigned to the content through auto-classification.
Recommendations
Stochastic Games with General Payoff Functions
We consider multiplayer stochastic games with finitely many players and actions, and countably many states, in which the payoff of each player is a bounded and Borel-measurable function of the infinite play. By using a generalization of the technique of ...
Correlation and unmediated cheap talk in repeated games with imperfect monitoring
AbstractThis paper studies n-player $$(n\ge 3)$$ undiscounted repeated games with imperfect monitoring. We prove that all uniform communication equilibrium payoffs of a repeated game can be obtained as Nash equilibrium payoffs of the game extended by ...
Cheap Talk Under Partial Conflicts: A Dynamic Analysis of Pragmatic Meaning
New Frontiers in Artificial IntelligenceAbstractIn natural language, meanings of words often deviate from their literal meanings under pragmatic reasoning. As is shown in game-theoretical pragmatics, when players do not have aligned benefits, communication with non-literal meaning is even more ...
Comments
Please enable JavaScript to view thecomments powered by Disqus.Information & Contributors
Information
Published In
Oct 2019
595 pages
ISBN:978-3-030-32429-2
DOI:10.1007/978-3-030-32430-8
© Springer Nature Switzerland AG 2019.
Publisher
Springer-Verlag
Berlin, Heidelberg
Publication History
Published: 30 October 2019
Author Tags
Qualifiers
- Article
Contributors
Other Metrics
Bibliometrics & Citations
Bibliometrics
Article Metrics
- 0Total Citations
- 0Total Downloads
- Downloads (Last 12 months)0
- Downloads (Last 6 weeks)0
Reflects downloads up to 25 Nov 2024
Other Metrics
Citations
View Options
View options
Login options
Check if you have access through your login credentials or your institution to get full access on this article.
Sign in