A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
A general reinforcement learning algorithm that masters chess, shogi, and Go through self-play
复制标题
DOI：
10.1126/science.aar6404
复制
发表时间：
2018-12-07
期刊：
影响因子：
56.9
通讯作者：
Hassabis, Demis
中科院分区：
文献类型：
--
作者：
Silver, David;Hubert, Thomas;Hassabis, Demis
国际象棋游戏是人工智能史上最长的研究领域。最强大的程序是基于复杂的搜索技术，特定领域的适应和手工制作的评估功能的结合，这些功能已被人类专家精炼了几十年。相比之下，Alphago Zero计划最近通过从自我玩法中学习而在GO的游戏中取得了超人的表现。在本文中，我们将这种方法推广到单个alphazero算法中，该算法可以在许多具有挑战性的游戏中实现超人性能。从随机游戏开始，除了游戏规则之外，没有任何领域知识，Alphazero令人信服地击败了国际象棋和Shogi（日本国际象棋）游戏中的世界冠军计划，也击败了GO。
The game of chess is the longest-studied domain in the history of artificial intelligence. The strongest programs are based on a combination of sophisticated search techniques, domain-specific adaptations, and handcrafted evaluation functions that have been refined by human experts over several decades. By contrast, the AlphaGo Zero program recently achieved superhuman performance in the game of Go by reinforcement learning from self-play. In this paper, we generalize this approach into a single AlphaZero algorithm that can achieve superhuman performance in many challenging games. Starting from random play and given no domain knowledge except the game rules, AlphaZero convincingly defeated a world champion program in the games of chess and shogi (Japanese chess), as well as Go.