论文信息 - Learning and Exploitation Do Not Conflict Under Minimax Optimality

Learning and Exploitation Do Not Conflict Under Minimax Optimality

We show that adaptive real time dynamic programming extended with the action selection strategy which chooses the best action according to the latest estimate of the cost function yields asymptotically optimal policies within finite time under the minimax optimality criterion. From this it follows that learning and exploitation do not conflict under this special optimality criterion. We relate this result to learning optimal strategies in repeated two-player zero-sum deterministic games.

Csaba Szepesvári | Csaba Szepesvari

[1] C. Stein. A Two-Sample Test for a Linear Hypothesis Whose Power is Independent of the Variance , 1945 .

[2] L. Shapley,et al. Stochastic Games* , 1953, Proceedings of the National Academy of Sciences.

[3] Onésimo Hernández-Lerma,et al. Controlled Markov Processes , 1965 .

[4] Richard E. Korf,et al. Real-Time Heuristic Search , 1990, Artif. Intell..

[5] Sebastian Thrun,et al. Learning to Play the Game of Chess , 1994, NIPS.

[6] Gerald Tesauro,et al. Temporal Difference Learning and TD-Gammon , 1995, J. Int. Comput. Games Assoc..

[7] Gerald Tesauro,et al. Temporal difference learning and TD-Gammon , 1995, CACM.

[8] Andrew G. Barto,et al. Learning to Act Using Real-Time Dynamic Programming , 1995, Artif. Intell..

[9] Csaba Szepesvári,et al. A Generalized Reinforcement-Learning Model: Convergence and Applications , 1996, ICML.

[10] Csaba Szepesv. Certainty Equivalence Policies Are Self-optimizing under Minimax Optimality Certainty Equivalent Policies Are Self-optimizing under Minimax Optimality , 1996 .

[11] epetivari Ctiaba Sz. Some basic facts concerning minimax sequential decision processes , 1996 .

[12] Csaba Szepesv Ari,et al. Generalized Markov Decision Processes: Dynamic-programming and Reinforcement-learning Algorithms , 1996 .