王宇新

个人信息Personal Information

副教授

硕士生导师

性别:男

毕业院校:大连理工大学

学位:博士

所在单位:计算机科学与技术学院

办公地点:创新园大厦A0827

联系方式:18640987378

电子邮箱:wyx@dlut.edu.cn

扫描关注

论文成果

当前位置: 中文主页 >> 科学研究 >> 论文成果

Research on actor-critic reinforcement learning in RoboCup

点击次数:

论文类型:会议论文

发表时间:2006-06-21

收录刊物:EI、CPCI-S、Scopus

卷号:2

页面范围:205-205

关键字:reinforcement learning; MAS; actor-critic; RoboCup; function approximation

摘要:Actor-Critic method combines the fast convergence of value-based (Critic) and directivity on search of policy gradient (Actor). It is suitable for solving the problems with large state space. In this paper, the Actor Critic method with the tile-coding linear function approximation is analysed and applied to a RoboCup simulation subtask named "Soccer Keepaway". The experiments on Soccer Keepaway show that the policy learned by Actor-Critic method is better than policies from value-based Sarsa(lambda) and benchmarks.