| 日時(Date) |
2024年11月25日 (月) / November 25th, 2024 3限 (13:30--15:00) / 3rd period (13:30--15:00) |
|---|---|
| 場所(Location) | 中講義室（L2） |
| 司会(Chair) | Takamitsu Matsubara |
| 講演者(Presenter) | Yunduan Cui (Shenzhen Institutes of Advanced Technology, Chinese Academy of Sciences (SIAT).) |
| 題目(Title) | Practical Reinforcement Learning for Engineering Applications |
| 概要(Abstract) | Applying reinforcement learning to real-world scenarios presents significant challenges, particularly in sampling efficiency and robustness against external disturbances. This lecture will introduce our research direction: practical reinforcement learning approaches for engineering applications. We aim to enhance reinforcement learning's efficiency during training and its resilience to uncertain disturbances by leveraging relative entropy constraints and probabilistic models. We've successfully validated this approach across diverse fields, including robotic dexterous hand, unmanned marine systems, and process control. |
| 講演言語(Language) | 英語 (English) |
| 講演者紹介(Introduction of Lecturer) | Yunduan Cui received his Ph.D. from Nara Institute of Science and Technology, Japan in 2017. From 2017 to 2020, he worked as a postdoctoral researcher and specially appointed assistant professor at Nara Institute of Science and Technology. In 2020, he joined Shenzhen Institute of Advanced Technology, Chinese Academy of Sciences as an associate professor. He has received the Best Oral Paper Award at IEEE-RAS International Conference on Humanoid Robots, the Annual Best Paper Award from Japanese Neural Network Society, and SICE Young Author Award for IROS. His research focuses on AI-driven automation systems, particularly the engineering applications of reinforcement learning technologies. |