论文信息 - Multiple Tree for Partially Observable Monte-Carlo Tree Search

Multiple Tree for Partially Observable Monte-Carlo Tree Search

We propose an algorithm for computing approximate Nash equilibria of partially observable games using Monte-Carlo tree search based on recent bandit methods. We obtain experimental results for the game of phantom tic-tac-toe, showing that strong strategies can be efficiently computed by our algorithm.

David Auger | D. Auger

[1] V. S. Subrahmanian,et al. Game-Tree Search with Combinatorially Large Belief States , 2005, IJCAI.

[2] Guillaume Chaslot,et al. A Comparison of Monte-Carlo Methods for Phantom Go , 2007 .

[3] Rémi Coulom,et al. Efficient Selectivity and Backup Operators in Monte-Carlo Tree Search , 2006, Computers and Games.

[4] Peter Auer,et al. The Nonstochastic Multiarmed Bandit Problem , 2002, SIAM J. Comput..

[5] Michael H. Bowling,et al. Regret Minimization in Games with Incomplete Information , 2007, NIPS.

[6] Tristan Cazenave,et al. A Phantom-Go Program , 2006, ACG.

[7] Gábor Lugosi,et al. Prediction, learning, and games , 2006 .

[8] David H. Li,et al. Kriegspiel: Chess Under Uncertainty , 1994 .

[9] M. Dufwenberg. Game theory. , 2011, Wiley interdisciplinary reviews. Cognitive science.

[10] Csaba Szepesvári,et al. Bandit Based Monte-Carlo Planning , 2006, ECML.

[11] Jean-Yves Audibert,et al. Minimax Policies for Adversarial and Stochastic Bandits. , 2009, COLT 2009.

[12] O. H. Brownlee,et al. ACTIVITY ANALYSIS OF PRODUCTION AND ALLOCATION , 1952 .