ShanghaiTech University Knowledge Management System
| LIGS: LEARNABLE INTRINSIC-REWARD GENERATION SELECTION FOR MULTI-AGENT LEARNING | |
| 2022-04 | |
| 会议录名称 | THE TENTH INTERNATIONAL CONFERENCE ON LEARNING REPRESENTATIONS |
| 发表状态 | 正式接收 |
| DOI | arxiv-2112.02618 |
| 摘要 | Efficient exploration is important for reinforcement learners to achieve high rewards. In multi-agent systems, coordinated exploration and behaviour is critical for agents to jointly achieve optimal outcomes. In this paper, we introduce a new general framework for improving coordination and performance of multi-agent reinforcement learners (MARL). Our framework, named Learnable Intrinsic-Reward Generation Selection algorithm (LIGS) introduces an adaptive learner, Generator that observes the agents and learns to construct intrinsic rewards online that coordinate the agents' joint exploration and joint behaviour. Using a novel combination of MARL and switching controls, LIGS determines the best states to learn to add intrinsic rewards which leads to a highly efficient learning process. LIGS can subdivide complex tasks making them easier to solve and enables systems of MARL agents to quickly solve environments with sparse rewards. LIGS can seamlessly adopt existing MARL algorithms and, our theory shows that it ensures convergence to policies that deliver higher system performance. We demonstrate its superior performance in challenging tasks in Foraging and StarCraft II. © 2022 ICLR 2022 - 10th International Conference on Learning Representationss. All rights reserved. |
| 会议录编者/会议主办者 | ByteDance ; et al. ; Meta AI ; Microsoft ; Qualcomm ; Sea Al Lab |
| 关键词 | Fertilizers Learning systems Reinforcement Reinforcement learning Coordinated behavior Coordinated exploration Intrinsic rewards Joint behavior Learn+ Learner control Multi-agent learning Multi-agent reinforcement Performance Selection algorithm |
| 会议名称 | 10th International Conference on Learning Representations, ICLR 2022 |
| 会议地点 | Virtual, Online |
| 会议日期 | April 25, 2022 - April 29, 2022 |
| 收录类别 | EI |
| 语种 | 英语 |
| 出版者 | International Conference on Learning Representations, ICLR |
| EI入藏号 | 暂无 |
| EI主题词 | Multi agent systems |
| EI分类号 | 723.4 Artificial Intelligence - 804 Chemical Products Generally - 821.2 Agricultural Chemicals - 951 Materials Science |
| 原始文献类型 | Conference article (CA) |
| 文献类型 | 会议论文 |
| 条目标识符 | https://kms.shanghaitech.edu.cn/handle/2MSLDSTB/183471 |
| 专题 | 信息科学与技术学院_硕士生 信息科学与技术学院_本科生 |
| 通讯作者 | Taher Jafferjee |
| 作者单位 | 1.ShanghaiTech University 2.Huawei Technologies 3.Imperial College London 4.Institute for AI, Peking University & BIGAI 5.University College London |
| 第一作者单位 | 上海科技大学 |
| 通讯作者单位 | 上海科技大学 |
| 第一作者的第一单位 | 上海科技大学 |
| 推荐引用方式 GB/T 7714 | David Mguni,Taher Jafferjee,Jianhong Wang,et al. LIGS: LEARNABLE INTRINSIC-REWARD GENERATION SELECTION FOR MULTI-AGENT LEARNING[C]//ByteDance, et al., Meta AI, Microsoft, Qualcomm, Sea Al Lab:International Conference on Learning Representations, ICLR,2022. |
| 条目包含的文件 | ||||||
| 文件名称/大小 | 文献类型 | 版本类型 | 开放类型 | 使用许可 | ||
修改评论
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。