在无知识条件下学习玩双人完全信息博弈
Learning to Play Two-Player Perfect-Information Games without Knowledge
AI总结：
本文提出树自举推广、无界深度极小极大修改、强化启发式和新动作选择分布等技术，通过无知识自对弈强化学习训练 Hex 程序，在 11 和 13 路棋盘上超越 Mohex 3HNN。
摘要：
本文提出了几种通过强化学习来学习博弈状态评估函数的技术。第一种是树自举（树学习）的推广：它被适配到基于非线性函数的无知识强化学习情境中。采用这种技术，在强化学习过程中不会丢失任何信息。第二种是对无界深度极小极大的修改，将最佳动作序列扩展到终局状态。这种修改后的搜索旨在学习过程中使用。第三种是用强化启发式替代经典的博弈收益（+1 / -1）。我们研究了特定的强化启发式，例如：快速获胜与缓慢失败；得分；机动性或存在性。第四种是一种新的动作选择分布。所进行的实验表明，这些技术提高了博弈水平。最后，我们应用这些不同的技术来设计 Hex 游戏（尺寸 11 和 13）的程序玩家，通过无知识的自对弈强化学习超越了 Mohex 3HNN 的水平。
英文摘要：
In this paper, several techniques for learning game state evaluation functions by reinforcement are proposed. The first is a generalization of tree bootstrapping (tree learning): it is adapted to the context of reinforcement learning without knowledge based on non-linear functions. With this technique, no information is lost during the reinforcement learning process. The second is a modification of minimax with unbounded depth extending the best sequences of actions to the terminal states. This modified search is intended to be used during the learning process. The third is to replace the classic gain of a game (+1 / -1) with a reinforcement heuristic. We study particular reinforcement heuristics such as: quick wins and slow defeats ; scoring ; mobility or presence. The four is a new action selection distribution. The conducted experiments suggest that these techniques improve the level of play. Finally, we apply these different techniques to design program-players to the game of Hex (size 11 and 13) surpassing the level of Mohex 3HNN with reinforcement learning from self-play without knowledge.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
