Optimal Correlated Equilibria in General-Sum Extensive-Form Games: Fixed-Parameter Algorithms, Hardness, and Two-Sided Column-Generation

被引:0
作者
Zhang, Brian Hu [1 ]
Farina, Gabriele [2 ]
Celli, Andrea [3 ]
Sandholm, Tuomas [1 ,4 ,5 ,6 ]
机构
[1] Carnegie Mellon Univ, Pittsburgh, PA 15232 USA
[2] MIT, Cambridge, MA 02139 USA
[3] Bocconi Univ, I-20136 Milan, MI, Italy
[4] Strategy Robot Inc, Pittsburgh, PA 15232 USA
[5] StrategicMachine Inc, Pittsburgh, PA 15232 USA
[6] Optimized Markets Inc, Pittsburgh, PA 15232 USA
基金
美国国家卫生研究院; 美国国家科学基金会;
关键词
equilibrium computation; extensive-form games; correlated equilibrium; COMPUTATION;
D O I
10.1287/moor.2022.0226
中图分类号
C93 [管理学]; O22 [运筹学];
学科分类号
070105 ; 12 ; 1201 ; 1202 ; 120202 ;
摘要
We study the problem of finding optimal correlated equilibria of various sorts in extensive-form games: normal-form coarse correlated equilibrium (NFCCE), extensive-form coarse correlated equilibrium (EFCCE), and extensive-form correlated equilibrium (EFCE). We make two primary contributions. First, we introduce a new algorithm for computing optimal equilibria in all three notions. Its runtime depends exponentially only on a parameter related to the information structure of the game. We also prove a fundamental complexity gap between NFCCE and the other two concepts. Second, we propose a two-sided column generation approach for use when the runtime or memory usage of the previous algorithm is prohibitive. Our algorithm improves upon an earlier one-sided approach by means of a new decomposition of correlated strategies which allows players to reoptimize their sequence-form strategies with respect to correlation plans which were previously added to the support. Experiments show that our techniques outperform the prior state of the art for computing optimal general-sum correlated equilibria.
引用
收藏
页数:33
相关论文
共 43 条
[1]  
Aumann R.J., 1974, J MATH ECON, V1, P67, DOI [10.1016/0304-4068(74)90037-8, DOI 10.1016/0304-4068(74)90037-8]
[2]   Superhuman AI for multiplayer poker [J].
Brown, Noam ;
Sandholm, Tuomas .
SCIENCE, 2019, 365 (6456) :885-+
[3]   Superhuman AI for heads-up no-limit poker: Libratus beats top professionals [J].
Brown, Noam ;
Sandholm, Tuomas .
SCIENCE, 2018, 359 (6374) :418-+
[4]  
Carminati L, 2022, PR MACH LEARN RES
[5]  
Celli A, 2019, Computing Optimal Ex Ante Correlated Equilibria in Two-Player Sequential Games
[6]  
Celli A, 2018, AAAI CONF ARTIF INTE, P965
[7]  
Celli Andrea, 2020, ADV NEURAL INFORM PR, V33
[8]   Tight lower bounds for certain parameterized NP-hard problems [J].
Chen, J ;
Chor, B ;
Fellows, M ;
Huang, XZ ;
Juedes, D ;
Kanj, IA ;
Xia, G .
INFORMATION AND COMPUTATION, 2005, 201 (02) :216-231
[9]   Settling the Complexity of Computing Two-Player Nash Equilibria [J].
Chen, Xi ;
Deng, Xiaotie ;
Teng, Shang-Hua .
JOURNAL OF THE ACM, 2009, 56 (03)
[10]   On the NP-completeness of finding an optimal strategy in games with common payoffs [J].
Chu, F ;
Halpern, J .
INTERNATIONAL JOURNAL OF GAME THEORY, 2001, 30 (01) :99-106