Presentation 2007-03-16
A computational study on motor skill aquisition
Rintaro TAGUCHI, Fumihiko ISHIDA, Yutaka SAKAGUCHI,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Although humans can acquire various motor skills by themselves, appropriate instructions by the expert instructors provides strong advantages for efficient learning, especially when we try to learn complex skills like skiing. In the present study, the authors built a reinforcement learning model which received "indirect reward", corresponding to such external instructions, in addition to the conventional reward telling the success/failure of the task, and examined how indirect reward affected the learning process through a computer-simulated experiment. The result shows that appropriate indirect reward assisted to build a superior internal model of the environment.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Motor skill / Reward / Reinforcement learning / Internal model
Paper # NC2006-190
Date of Issue

Conference Information
Committee NC
Conference Date 2007/3/9(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 Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) A computational study on motor skill aquisition
Sub Title (in English)
Keyword(1) Motor skill
Keyword(2) Reward
Keyword(3) Reinforcement learning
Keyword(4) Internal model
1st Author's Name Rintaro TAGUCHI
1st Author's Affiliation Graduate School of Information Systems, University of Electro-Communications()
2nd Author's Name Fumihiko ISHIDA
2nd Author's Affiliation Graduate School of Information Systems, University of Electro-Communications
3rd Author's Name Yutaka SAKAGUCHI
3rd Author's Affiliation Graduate School of Information Systems, University of Electro-Communications
Date 2007-03-16
Paper # NC2006-190
Volume (vol) vol.106
Number (no) 590
Page pp.pp.-
#Pages 6
Date of Issue