arXivDaily arXiv每日学术速递 周一至周五更新
arXiv周末暂无论文更新,休息一下吧,周末愉快~~
arXiv 2602.23318cs.AI
收藏

内存受限环境中的通用快速动作价值估计

Generalized Rapid Action Value Estimation in Memory-Constrained Environments

ICTEAM, UCLouvain(ICTEAM,UCLouvain大学);LAMSADE, Université Paris-Dauphine(LAMSADE,巴黎-dauphine大学)

Aloïs Rautureau, Tristan Cazenave, Éric Piette

AI总结:

本文提出GRAVE2、GRAVER和GRAVER2算法,通过两层搜索、节点回收等方法,在内存受限环境下提升游戏性能。

摘要:

通用快速动作价值估计(GRAVE)已被证明是蒙特卡洛树搜索(MCTS)家族算法中用于一般游戏编程(GGP)的强大变种。然而,其在每个节点存储额外的胜负/访问统计信息的依赖性使其在内存受限的环境中使用变得不切实际,从而限制了其在实践中的应用。在本文中,我们引入了GRAVE2、GRAVER和GRAVER2算法,分别通过两层搜索、节点回收和两者的结合扩展了GRAVE。我们证明这些改进使存储的节点数量大幅减少,同时保持与GRAVE相媲美的游戏实力。

英文摘要:

Generalized Rapid Action Value Estimation (GRAVE) has been shown to be a strong variant within the Monte-Carlo Tree Search (MCTS) family of algorithms for General Game Playing (GGP). However, its reliance on storing additional win/visit statistics at each node makes its use impractical in memory-constrained environments, thereby limiting its applicability in practice. In this paper, we introduce the GRAVE2, GRAVER and GRAVER2 algorithms, which extend GRAVE through two-level search, node recycling, and a combination of both techniques, respectively. We show that these enhancements enable a drastic reduction in the number of stored nodes while matching the playing strength of GRAVE.

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。

↑