Strategy Acquisition for the Game "Othello" Based on Reinforcement Learning

スポンサーリンク

概要

• 論文の詳細を見る

• This article discusses automatic strategy acquisition for the game "Othello" based on a reinforcement learning scheme. In our approach, a computer player, which initially knows only the game rules, becomes stronger after playing several thousands of games against another player. In each game, the computer player refines the evaluation function for the game state, which is achieved by min-max reinforcement learning (MMRL). MMRL is a simple learning scheme that uses the minmax strategy. Since the state space of Othello is huge, we employ a normalized Gaussian network (NGnet) to represent the evaluation function. As a result, the computer player becomes strong enough to beat a player employing a heuristic strategy. This article experimentally shows that MMRL is better than TD(0) and also shows that the NGnet is better than a multi-layered perceptron, in our Othello task.

• 社団法人電子情報通信学会の論文
• 1999-12-25

著者

• Ishii Shin

  Nara Inst. Of Sci. And Technol. 8916-5 Takayama-cho Ikoma Nara 630-0192 Jpn

• Yoshioka Taku

  Nara Institute Of Science And Technology

• ITO Minoru

  Nara Institute of Science and Technology

関連論文

• On-Line Learning Methods for Gaussion Processes
• Reinforcement learning for a biped robot based on a CPG-actor-critic method
• Stochastic resonance with differential code in feedforward network with intra-layer random connections
• Strategy Acquisition for the Game "Othello" Based on Reinforcement Learning
• Control of exploitation-exploration meta-parameter in reinforcement learning
• Bayesian representation learning in the cortex regulated by acetylcholine
• A New Weight Function that Improves the Resolution in Radial Direction in Direct Dipole Estimation for Poisson Equation

スポンサーリンク