Presentation 2000/5/18
Shoot Behavior Acquisition of Autonomous Robot Using Reinforcement Learning
Akimoto Takashi, Miyauchi Arata, Ishikawa Tomoo,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this article, we examine Q-learning which considered learning time. In Q-learning method, each subtask requires its own Q-table. When the number of subtasks increases, the ratio between the learning time at each subtask to total learning time decreases, so we need much time to finish learning of all subtasks. So, we propose the learning method to handle two subtasks as one and use our Q-table. We applied our method into shoot behavior acquisition of soccer robot. As the result of experiments, indicates the effectiveness of our method(using the same learning result in different two subtasks).
Keyword(in Japanese) (See Japanese page)
Keyword(in English) real robot / reinforcement learning / on-line learning / soccer robot
Paper # AI2000-1
Date of Issue

Conference Information
Committee AI
Conference Date 2000/5/18(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Artificial Intelligence and Knowledge-Based Processing (AI)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Shoot Behavior Acquisition of Autonomous Robot Using Reinforcement Learning
Sub Title (in English)
Keyword(1) real robot
Keyword(2) reinforcement learning
Keyword(3) on-line learning
Keyword(4) soccer robot
1st Author's Name Akimoto Takashi
1st Author's Affiliation Graduate School of Electrical Engineering Musashi Institute of Technology()
2nd Author's Name Miyauchi Arata
2nd Author's Affiliation Graduate School of Electrical Engineering Musashi Institute of Technology
3rd Author's Name Ishikawa Tomoo
3rd Author's Affiliation Graduate School of Electrical Engineering Musashi Institute of Technology
Date 2000/5/18
Paper # AI2000-1
Volume (vol) vol.100
Number (no) 88
Page pp.pp.-
#Pages 6
Date of Issue