arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 2005.11335cs.LGcs.AI
收藏

基于蒙特卡罗树搜索的单代理策略迭代优化

Single-Agent Optimization Through Policy Iteration Using Monte-Carlo Tree Search

University of Alberta(阿尔伯塔大学)

Arta Seify, Michael Buro

摘要:

蒙特卡罗树搜索(MCTS)和深度强化学习的结合是两人完全信息博弈中的最新技术。在本文中,我们描述了一种搜索算法,使用MCTS的一个变体,我们增强了1)一种新的动作值规范化机制的游戏与潜在的无限奖励(这是在许多优化问题的情况下),2)定义一个虚拟的损失函数,使有效的搜索并行化,和3)的政策网络,训练的几代人的自我发挥,以指导搜索。我们衡量我们的方法在“SameGame”-一个流行的单人游戏测试域的有效性。我们的实验结果表明,我们的方法优于基线算法在几个电路板尺寸。此外,它与最先进的搜索算法在公共位置集上具有竞争力。

英文摘要:

The combination of Monte-Carlo Tree Search (MCTS) and deep reinforcement learning is state-of-the-art in two-player perfect-information games. In this paper, we describe a search algorithm that uses a variant of MCTS which we enhanced by 1) a novel action value normalization mechanism for games with potentially unbounded rewards (which is the case in many optimization problems), 2) defining a virtual loss function that enables effective search parallelization, and 3) a policy network, trained by generations of self-play, to guide the search. We gauge the effectiveness of our method in "SameGame"---a popular single-player test domain. Our experimental results indicate that our method outperforms baseline algorithms on several board sizes. Additionally, it is competitive with state-of-the-art search algorithms on a public set of positions.

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。

↑