Developments on Product Propagation
Developments on Product Propagation
复制标题
产品传播的进展
DOI：
--
复制
发表时间：
2013
期刊：
影响因子：
--
通讯作者：
T. Cazenave
中科院分区：
文献类型：
--
作者：
Abdallah Saffidine;T. Cazenave
乘积传播（Product Propagation，PP）是一种为抽象的两人游戏备份概率评估的算法。结果表明，pp可以像证明数搜索（pns）一样有效地解决Go问题。在本文中，我们展示了三个领域，对于通用的非优化版本，pp比以前已知的解决游戏的算法表现更好（见论文中的细微差别）。比较的方法包括alpha-beta搜索、pns和蒙特-卡罗树搜索。我们还扩展了pp，以处理其内存消耗和提高其求解时间。
Product Propagation (pp) is an algorithm to backup probabilistic evaluations for abstract two-player games. It was shown that pp could solve Go problems as efficiently as Proof Number Search (pns). In this paper, we exhibit three domains where, for generic non-optimized versions, pp performs better (see the nuances in the paper) than previously known algorithms for solving games. The compared approaches include alpha-beta search, pns, and Monte-Carlo Tree Search. We also extend pp to deal with its memory consumption and to improve its solving time.