Title :
Rule abstraction and transfer in reinforcement learning by decision tree
Author :
Min Wu ; Yamashita, Atsushi ; Asama, Hajime
Author_Institution :
Dept. of Precision Eng., Univ. of Tokyo, Tokyo, Japan
Abstract :
Reinforcement learning agents store their knowledge such as state-action value in look-up tables. However, loop-up table requires large memory space when number of states become large. Learning from look-up table is tabularasa therefore is very slow. To overcome this disadvantage, generalization methods are used to abstract knowledge. In this paper, decision tree technology is used to enable the agent to represent abstract knowledge in rule from during learning progress and form rule base for each individual task.
Keywords :
decision trees; learning (artificial intelligence); table lookup; abstract knowledge; decision tree; decision tree technology; generalization methods; large memory space; look-up tables; reinforcement learning; tabularasa; Abstracts; Decision trees; Educational institutions; Learning; Learning systems; Robots; Table lookup;
Conference_Titel :
System Integration (SII), 2012 IEEE/SICE International Symposium on
Conference_Location :
Fukuoka
Print_ISBN :
978-1-4673-1496-1
DOI :
10.1109/SII.2012.6427332