【摘要】 强化学习是一种重要的机器学习方法,随着计算机网络和分布式处理技术的飞速发展,多智能体系统中的分布式强化学习方法正受到越来越多的关注。论文将目前已有的各种分布式强化学习方法总结为中央强化学习、独立强化学习、群体强化学习、社会强化学习四类,然后探讨了这四类分布式强化学习方法的体系结构框架,并给出了这四类分布式强化学习方法的形式化定义。更多还原

【Abstract】 Reinforcement learning is a kind of important machine learning algorithm.With the development of computer network technique and distributed process technique,distributed reinforcement learning algorithms in multi-agent systems are receiving more and more attentions and interests.In this paper,all the present kinds of distributed reinforcement learning algorithms are classified into four categories,which are respectively named as Reinforcement Learning Centrally(RLC),Reinforcement Learning Individually(RLI ),Reinforcement Learning in Groups(RLG)and Reinforcement Learning Socially(RLS).Whereafter,the architectures of the four categories are analyzed and discussed,and their formal definitions are presented.更多还原