DocumentCode :
3180436
Title :
Heterogeneous and Hierarchical Cooperative Learning via Combining Decision Trees
Author :
Asadpour, Masoud ; Ahmadabadi, Majid Nili ; Siegwart, Roland
Author_Institution :
Autonomous Syst. Lab., Ecole Polytech. Fed. de Lausanne
fYear :
2006
fDate :
9-15 Oct. 2006
Firstpage :
2684
Lastpage :
2690
Abstract :
Decision trees, being human readable and hierarchically structured, provide a suitable mean to derive state-space abstraction and simplify the inclusion of the available knowledge for a reinforcement learning (RL) agent. In this paper, we address two approaches to combine and purify the available knowledge in the abstraction trees, stored among different RL agents in a multi-agent system, or among the decision trees learned by the same agent using different methods. Simulation results in nondeterministic football learning task provide strong evidences for enhancement in convergence rate and policy performance
Keywords :
decision trees; hierarchical systems; learning (artificial intelligence); multi-agent systems; state-space methods; abstraction trees; decision trees; hierarchical cooperative learning; multi-agent system; nondeterministic football learning task; reinforcement learning; state-space abstraction; Control systems; Decision trees; Humans; Intelligent agent; Intelligent control; Intelligent robots; Intelligent structures; Learning systems; Multiagent systems; Process control;
fLanguage :
English
Publisher :
ieee
Conference_Titel :
Intelligent Robots and Systems, 2006 IEEE/RSJ International Conference on
Conference_Location :
Beijing
Print_ISBN :
1-4244-0258-1
Electronic_ISBN :
1-4244-0259-X
Type :
conf
DOI :
10.1109/IROS.2006.281990
Filename :
4058796
Link To Document :
بازگشت