Knowledge Commons of Institute of Automation,CAS
Centralized Cooperative Exploration Policy for Continuous Control Tasks | |
Chao Li1; Chen Gong1; Qiang He2; Xinwen Hou1; Yu Liu1 | |
2023-05 | |
会议名称 | the 2023 International Conference on Autonomous Agents and Multiagent Systems |
会议录名称 | Proceedings of the 2023 International Conference on Autonomous Agents and Multiagent Systems |
页码 | 2454–2456 |
会议日期 | May 29–June 2, 2023 |
会议地点 | London, United Kingdom |
出版地 | Richland, SC |
出版者 | International Foundation for Autonomous Agents and Multiagent Systems |
摘要 | Despite recent works making great progress in continuous control tasks, exploration in these tasks has remained insufficiently investigated. This paper proposes CCEP (C entralized C ooperative E xploration P olicy), which utilizes estimation biases of value functions to contribute to the exploration capacity. CCEP keeps two value functions initialized with different parameters, and generates diverse policies with multiple exploration styles from a pair of value functions. In addition, a centralized policy framework ensures that CCEP achieves message delivery between multiple policies, furthermore contributing to exploring the environment cooperatively. Extensive experimental results demonstrate that CCEP achieves higher exploration capacity. Empirical analysis shows diverse exploration styles in the learned policies by CCEP, reaping benefits in more exploration regions. Besides, the exploration capabilities of CCEP have been demonstrated to outperform current state-of-the-art methods on multiple continuous control tasks. |
关键词 | continuous control tasks cooperative exploration |
DOI | 10.5555/3545946.3598965 |
收录类别 | EI |
语种 | 英语 |
是否为代表性论文 | 否 |
七大方向——子方向分类 | 强化与进化学习 |
国重实验室规划方向分类 | 认知决策知识体系 |
是否有论文关联数据集需要存交 | 否 |
引用统计 | |
文献类型 | 会议论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/56696 |
专题 | 多模态人工智能系统全国重点实验室_机器人理论与应用 |
通讯作者 | Xinwen Hou; Yu Liu |
作者单位 | 1.Institute of Automation, Chinese Academy of Sciences, Beijing, China 2.University of Tubingen, Tubingen, Germany |
第一作者单位 | 中国科学院自动化研究所 |
通讯作者单位 | 中国科学院自动化研究所 |
推荐引用方式 GB/T 7714 | Chao Li,Chen Gong,Qiang He,et al. Centralized Cooperative Exploration Policy for Continuous Control Tasks[C]. Richland, SC:International Foundation for Autonomous Agents and Multiagent Systems,2023:2454–2456. |
条目包含的文件 | 下载所有文件 | |||||
文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
p2454.pdf(2175KB) | 会议论文 | 开放获取 | CC BY-NC-SA | 浏览 下载 |
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