Proceedings of the Fuzzy System Symposium
31st Fuzzy System Symposium
Session ID : WE2-5
Conference information

main
Heterogeneous multi-agent simulation for Fuzzy Q-learning and Transfer learning by using the common behavior estimation
*Mutsumi IwasaNaoyuki Kubota
Author information
CONFERENCE PROCEEDINGS FREE ACCESS

Details
Abstract
n this paper, we analyse the acquired behavior in predation problem by heterogeneous multi-agent.We also clarify the relationship of the common behavior and individual behavior through difference of embodiment of each agent.It is shown that we can acquire continuous behavior in grid-like discrete space by using fuzzy Q-learning.Finally, we discuss the relevance between universal portion of the common behavior obtained by fuzzy Q-learning and basic behavior in transition learning.
Content from these authors
© 2015 Japan Society for Fuzzy Theory and Intelligent Informatics
Previous article Next article
feedback
Top