Using Modified UCT Algorithm Basing on Risk Estimation Methods in Imperfect Information Games
نویسندگان
چکیده
Risk dominance and payoff dominance strategy are two complementary parts of the game theory decision strategy. While payoff dominance is still the basic principle in perfect information, two player games, risk dominance has shown its advantages in imperfect information conditions. In this paper, we first review the related work in the area of estimation methods and the influence of risk factors on computing game equilibrium. Then a new algorithm, UCT-Risk is proposed in this paper, which is a modification of UCT (UCB apply to Trees) algorithm based on risk estimation methods. Finally, we implement the proposed algorithm in SiGuo game, a popular imperfect information game in China. The experimental result of the new algorithm shows it correctness and effectiveness.
منابع مشابه
A Modified UCT Algorithm Basd on Risk Estimation Methods
Risk dominance and payoff dominance strategy are two complementary parts of the game theory decision strategy. While payoff dominance is still the basic principle in perfect information, two player games, risk dominance has shown its advantages in imperfect information conditions. In this paper, we first review the related work in the area of estimation methods and the influence of risk factors...
متن کاملSmooth UCT Search in Computer Poker
Self-play Monte Carlo Tree Search (MCTS) has been successful in many perfect-information twoplayer games. Although these methods have been extended to imperfect-information games, so far they have not achieved the same level of practical success or theoretical convergence guarantees as competing methods. In this paper we introduce Smooth UCT, a variant of the established Upper Confidence Bounds...
متن کاملSelf-Play Monte-Carlo Tree Search in Computer Poker
Self-play reinforcement learning has proved to be successful in many perfect information two-player games. However, research carrying over its theoretical guarantees and practical success to games of imperfect information has been lacking. In this paper, we evaluate selfplay Monte-Carlo Tree Search (MCTS) in limit Texas Hold’em and Kuhn poker. We introduce a variant of the established UCB algor...
متن کاملEquilibrium Selection in Evolutionary Games with Imperfect Monitoring
In this paper we analyze players’ long-run behavior in evolutionary coordination games with imperfect monitoring in a large population. Players can observe signals corresponding to other players’ unseen actions and use the proposed simple or maximum likelihood estimation algorithm to extract information from the signals. In the simple learning process we find conditions for the risk-dominant an...
متن کاملBuilding Opponent Model in Imperfect Information Board Games
In imperfect information problems, board game is a class of special problem that differs from card games like poker. Several characters make it a valuable test bed for opponent modeling, which is one of the most difficult problems in artificial intelligence decision systems. In card games, opponent modeling has proved its importance on improving agents’ strength. In this paper, a method of buil...
متن کامل