Reinforcement Learning Algorithm for Solving RTDP with Variational Environment

Changming Yin, Lihua Xue, Minghui Hu, Liyun Li

College of Computer and Communicational Engineering, Changsha University of Science and Technology, 45 Chiling Road, Changsha, Hunan, 410076, China
http://www.csust.edu.cn

Abstract:

Reinforcement learning algorithm based-on solving RTDP does not have to evaluate the entire state space in order to deliver an optimal policy, and yet it can often obtain good policy. The DP always is employed to be a learning Agent's environment. In this paper, we propose a novel algorithm called V-RTDP to solve Agent learning problems with full observability. Here, we use a new DP framework in varying environment instead of a general one. Our experiment demonstrates that the algorithm is very efficient.

Keywords: Real-time Dynamic Programming, Variational Environment, Reinforcement Learning, Algorithm, Optimal Policy

( Full Paper, 636 KB)

Biographies:

	Changming Yin received B.Sc degree from Beijing Normal University, Ph.D from Shanghai University, China,in the year 1986 and 2006 respectively. He is presently a professor in the College of Computer and Communicational Engineering, Changsha University of Science and Technology. His research areas include intelligence control and simulations, intelligence computing and machine learning, etc.		Lihua Xue received B.Sc degree from Taiyuan Normal University, China,in the year 2005. She is presently a graduate in the College of Computer and Communicational Engineering, Changsha University of Science and Technology. Her research areas include artificial intelligence and machine learning, etc.
	Minghui Hu received B.Sc degree from Changsha University of Science and Technology, China,in the year 2005. He is presently a graduate in the College of Computer and Communicational Engineering, Changsha University of Science and Technology. His research areas include software engineering, intelligence computing and machine learning, etc.		Liyun Li received B.Sc degree from Huadong Normal University, Shanghai, China,in the year 1998. He is presently a graduate in the College of Computer and Communicational Engineering, Changsha University of Science and Technology. His research areas include intelligence control and simulations, etc.

BibTex:

@ARTICLE{P1120715001,

AUTHOR = {Changming Yin and Lihua Xue and Minghui Hu and Liyun Li},

TITLE = {Reinforcement Learning Algorithm for Solving RTDP with Variational Environment},

JOURNAL =  {ICGST  International Journal on Artificial Intelligence and Machine Learning, AIML},

YEAR = {2007},

VOLUME = {7},

ISSUE ={1},

PAGES={17--21}

}

( Full Paper, 636 KB)