Author:
Han Yuntao,Zhou Qibin,Duan Fuqing
Abstract
AbstractThe digital curling game is a two-player zero-sum extensive game in a continuous action space. There are some challenging problems that are still not solved well, such as the uncertainty of strategy, the large game tree searching, and the use of large amounts of supervised data, etc. In this work, we combine NFSP and KR-UCT for digital curling games, where NFSP uses two adversary learning networks and can automatically produce supervised data, and KR-UCT can be used for large game tree searching in continuous action space. We propose two reward mechanisms to make reinforcement learning converge quickly. Experimental results validate the proposed method, and show the strategy model can reach the Nash equilibrium.
Publisher
Springer Science and Business Media LLC
Subject
General Earth and Planetary Sciences,General Environmental Science
Reference22 articles.
1. Roughgarden T (2010) Algorithmic game theory. Commun ACM 53(7):78–86
2. Wang FY, Zhang JJ, Zheng X et al (2016) Where does AlphaGo go: from church-turing thesis to AlphaGo thesis and beyond. IEEE CAA J Autom Sin 3(2):113–120
3. Zinkevich M, Johanson M, Bowling M et al (2007) Regret minimization in games with incomplete information. Adv Neural Inf Process Syst 20:1729–1736
4. Heinrich J, Silver D (2016) Deep reinforcement learning from self-play in imperfect-information games. arXiv preprint arXiv:1603.01121
5. Heinrich J, Lanctot M, Silver D (2015) Fictitious self-play in extensive-form games. In: International conference on machine learning. pp 805–813
Cited by
9 articles.
订阅此论文施引文献
订阅此论文施引文献，注册后可以免费订阅5篇论文的施引文献，订阅后可以查看论文全部施引文献