Presentation 2000/5/18
Automatic control of continuous systems based on on-line EM reinforcement learning
Yoshimoto Junichiro, Ishii Shin, Sato Masa-aki,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this report, we propose a new reinforcement learning(RL)method for continuous dynamical systems by using function approximation and stochastic learning. Our RL method has an architecture like the actor-critic model. The critic tries to approximate the Q-function, which is the expected future return for the current state-action pair. The actor tries to approximate a stochastic soft-max policy defined by the Q-function. The soft-max policy is more likely to select an action that has a higher Q-function value. The on-line EM algorithm is used to train the critic and the actor. We apply this method to two control problems. Computer simulations show that our method is able to acquire faurly good control in the two tasks after a few learning trials.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) reinforcement learning / actor-critic model / continuous dynamical system / stochastic model / EM algorithm
Paper # AI2000-5
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) Automatic control of continuous systems based on on-line EM reinforcement learning
Sub Title (in English)
Keyword(1) reinforcement learning
Keyword(2) actor-critic model
Keyword(3) continuous dynamical system
Keyword(4) stochastic model
Keyword(5) EM algorithm
1st Author's Name Yoshimoto Junichiro
1st Author's Affiliation Nara Institute of Science and Technology()
2nd Author's Name Ishii Shin
2nd Author's Affiliation Nara Institute of Science and Technology :CREST, Japan Science and Technology Corporation
3rd Author's Name Sato Masa-aki
3rd Author's Affiliation ATR International:CREST, Japan Science and Technology Corporation
Date 2000/5/18
Paper # AI2000-5
Volume (vol) vol.100
Number (no) 88
Page pp.pp.-
#Pages 8
Date of Issue