arXivDaily arXiv每日学术速递 周一至周五更新
arXiv 1707.06307cs.GT
收藏

强化学习产生迭代囚徒困境的优势策略

Reinforcement Learning Produces Dominant Strategies for the Iterated Prisoner's Dilemma

Marc Harper, Vincent Knight, Martin Jones, Georgios Koutsovoulos, Nikoleta E. Glynatsi, Owen Campbell

摘要:

我们目前的比赛结果和几个强大的策略,使用强化学习技术(进化和粒子群算法)创建的迭代囚徒困境。这些策略经过训练,可以在超过170个不同对手的语料库中表现良好,其中包括许多着名和经典的策略。所有经过训练的策略都能在标准锦标赛中战胜所有其他对手。经过训练的策略和一个特定的人工设计的策略在嘈杂的比赛中也表现最好。

英文摘要:

We present tournament results and several powerful strategies for the Iterated Prisoner's Dilemma created using reinforcement learning techniques (evolutionary and particle swarm algorithms). These strategies are trained to perform well against a corpus of over 170 distinct opponents, including many well-known and classic strategies. All the trained strategies win standard tournaments against the total collection of other opponents. The trained strategies and one particular human made designed strategy are the top performers in noisy tournaments also.

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。

↑