Effective strategies for complex skill real-time learning using reinforcement learning | |
Wei YZ(魏英姿); Zhao MY(赵明扬) | |
2003 | |
会议名称 | IEEE International Conference on Robotics, Intelligent Systems and Signal Processing |
会议日期 | October 8-13, 2003 |
会议地点 | Changsha, China |
页码 | 388-392 |
中文摘要 | Following the principle of human skill learning, robot acquiring skill is a process similar to human skill learning. Reinforcement learning is on-line actor critic method for robot to develop its skill. The reinforcement function has become the critical component for its effect of evaluating the action and guiding the learning process. A difference form of augmented reward function is considered carefully. In this paper we present a strategy for the task of complex skill learning. Automatic robot shaping policy is to dissolve the complex skill into a hierarchical learning process. Variable resolution discretization of input space is introduced to improve the generalization capability of CMAC-based RL. Conventional epsilon-greedy policy has the shortage of unnecessary randomization. Boltzmann distribution selection is also introduced to the balance of exploration and exploitation. We describe our ideas of reinforcement learning methods and also illustrate with an example the utility of method for learning skilled robot control on line. |
收录类别 | CPCI(ISTP) |
产权排序 | 1 |
会议主办者 | Chinese High tech Dev Program, Chinese Acad Sci, Chinese Soc Automat, IEEE Syst, Man & Cybernet Soc, Chinese Univ Hong Kong, Natl Univ Def Technol, IEEE Hong Kong RA & CS Joint Chapter, Joint Ctr Intelligen Sensing & Syst, Sch Elect Sci & Engn, Sch Mechatron Engn & Automat |
会议录 | 2003 IEEE INTERNATIONAL CONFERENCE ON ROBOTICS, INTELLIGENT SYSTEMS AND SIGNAL PROCESSING, VOLS 1 AND 2, PROCEEDINGS |
会议录出版者 | IEEE |
会议录出版地 | NEW YORK |
语种 | 英语 |
ISBN号 | 0-7803-7925-X |
WOS记录号 | WOS:000189506600068 |
内容类型 | 会议论文 |
源URL | [http://ir.sia.cn/handle/173321/8685] |
专题 | 沈阳自动化研究所_机器人学研究室 |
推荐引用方式 GB/T 7714 | Wei YZ,Zhao MY. Effective strategies for complex skill real-time learning using reinforcement learning[C]. 见:IEEE International Conference on Robotics, Intelligent Systems and Signal Processing. Changsha, China. October 8-13, 2003. |
个性服务 |
查看访问统计 |
相关权益政策 |
暂无数据 |
收藏/分享 |
除非特别说明,本系统中所有内容都受版权保护,并保留所有权利。
修改评论