Encoding Primitives Generation Policy Learning for Robotic Arm to Overcome Catastrophic Forgetting in Sequential Multi-tasks Learning | |
Xiong, Fangzhou1,2; Liu, Zhiyong1,2,3; Huang, Kaizhu4,5; Yang, Xu1,2; Qiao, Hong1,2,3; Amir Hussain6 | |
发表期刊 | Neural Networks |
2020-06 | |
期号 | 2020.06.003页码:12 |
摘要 | Continual learning, a widespread ability in people and animals, aims to learn and acquire new knowledge and skills continuously. Catastrophic forgetting usually occurs in continual learning when an agent attempts to learn different tasks sequentially without storing or accessing previous task information. Unfortunately, current learning systems, e.g., neural networks, are prone to deviate the weights learned in previous tasks after training new tasks, leading to catastrophic forgetting, especially in a sequential multi-tasks scenario. To address this problem, in this paper, we propose to overcome catastrophic forgetting with the focus on learning a series of robotic tasks sequentially. Particularly, a novel hierarchical neural network's framework called Encoding Primitives Generation Policy Learning (E-PGPL) is developed to enable continual learning with two components. By employing a variational autoencoder to project the original state space into a meaningful low-dimensional feature space, representative state primitives could be sampled to help learn corresponding policies for different tasks. In learning a new task, the feature space is required to be close to the previous ones so that previously learned tasks can be protected. Extensive experiments on several simulated robotic tasks demonstrate our method's efficacy to learn control policies for handling sequentially arriving multi-tasks, delivering improvement substantially over some other continual learning methods, especially for the tasks with more diversity. |
关键词 | Sequential multi-tasks learning, Continual learning, Catastrophic forgetting, Robotics |
收录类别 | SCI |
WOS记录号 | WOS:000555927200014 |
七大方向——子方向分类 | 强化与进化学习 |
引用统计 | |
文献类型 | 期刊论文 |
条目标识符 | http://ir.ia.ac.cn/handle/173211/39080 |
专题 | 复杂系统管理与控制国家重点实验室_机器人理论与应用 |
通讯作者 | Liu, Zhiyong |
作者单位 | 1.State Key Lab of Management and Control for Complex Systems, Institute of Automation, Chinese Academy of Science, 100190, Beijing, China 2.School of Artificial Intelligence, University of Chinese Academy of Sciences (UCAS), 100049, Beijing, China 3.CAS Centre for Excellence in Brain Science and Intelligence Technology, Chinese Academy of Sciences, 200031, Shanghai, China 4.Department of EEE, Xi'an Jiaotong-Liverpool University, 215123, Suzhou, China 5.Alibaba-Zhejiang University Joint Institute of Frontier Technologies, Hangzhou, China 6.Cyber and Big Data Research Laboratory, Edinburgh Napier University, Edinburgh EH11 4BN, U.K. |
推荐引用方式 GB/T 7714 | Xiong, Fangzhou,Liu, Zhiyong,Huang, Kaizhu,et al. Encoding Primitives Generation Policy Learning for Robotic Arm to Overcome Catastrophic Forgetting in Sequential Multi-tasks Learning[J]. Neural Networks,2020(2020.06.003):12. |
APA | Xiong, Fangzhou,Liu, Zhiyong,Huang, Kaizhu,Yang, Xu,Qiao, Hong,&Amir Hussain.(2020).Encoding Primitives Generation Policy Learning for Robotic Arm to Overcome Catastrophic Forgetting in Sequential Multi-tasks Learning.Neural Networks(2020.06.003),12. |
MLA | Xiong, Fangzhou,et al."Encoding Primitives Generation Policy Learning for Robotic Arm to Overcome Catastrophic Forgetting in Sequential Multi-tasks Learning".Neural Networks .2020.06.003(2020):12. |
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 | ||
Manuscript.pdf(646KB) | 期刊论文 | 作者接受稿 | 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