论文部分内容阅读
2016年3月以及2017年5月,在与李世石和柯洁两位最顶尖人类棋手的两次围棋人机大战中,Alpha Go分别以4:1和3:0的比分获胜。围棋界从对计算机围棋评价不高,到承认计算机已经彻底战胜了人类棋手,只有短短的一年多时间。文章介绍蒙特卡罗树形搜索、策略网络、价值网络、强化学习等围棋算法思想,回顾计算机围棋算法不断发展直至处理复杂的能力超过人类棋手的历程,并展望人工智能对围棋与社会的影响。
In March 2016 and May 2017, Alpha Go won 4: 1 and 3: 0 respectively in two go man-machine warfare against two of the top human chess players, Li Shishi and Ke Jie. Go community from the evaluation of computer Go is not high, to admit that the computer has completely defeated the human chess player, only a short period of more than a year. The article introduces the idea of Go algorithm such as tree search, strategy network, value network and reinforcement learning in Monte Carlo. It reviews the progress of computer Go algorithm until it deals with the complicated ability surpassing that of human chess player and looks forward to the influence of artificial intelligence on Go and society .