Agent Learning Instead of Behavior Implementation for Simulations
--- A Case Study Using Classifier Systems
F. Klügl, R. Hatko, and M. Butz. MATES 2008: Proc. 6th German Conf. on Multiagent System
Technologies, volume 5244 of LNCS, page 111--122. Kaiserslautern, Germany, Springer, (23-26 Sept. 2008)
DOI: 10.1007/978-3-540-87805-6_11
Abstract
Although multi-agent simulations are an intuitive way of conceptualizing
systems that consist of autonomous actors, a major problem is the
actual design of the agent behavior. In this contribution, we examine
the potential of using agent-based learning for implementing the
agent behavior. We enhanced SeSAm, a platform for agent-based simulation,
by replacing the usual rule-based agent architecture by XCS, a well-known
learning classifier system (LCS). The resulting model is tested using
a simple evacuation scenario. The results show that on the one hand
side plausible agent behavior could be learned. On the other hand
side, though, the results are quite brittle concerning the frame
of environmental feedback, perception and action modeling.
%0 Conference Paper
%1 Kluegl:2008:mates
%A Klügl, Franziska
%A Hatko, Reinhard
%A Butz, Martin V.
%B MATES 2008: Proc. 6th German Conf. on Multiagent System
Technologies
%C Kaiserslautern, Germany
%D 2008
%E Bergmann, Ralph
%E Lindemann, Gabriela
%E Kirn, Stefan
%E Pechoucek, Michal
%I Springer
%K imported thesis
%P 111--122
%R 10.1007/978-3-540-87805-6_11
%T Agent Learning Instead of Behavior Implementation for Simulations
--- A Case Study Using Classifier Systems
%V 5244
%X Although multi-agent simulations are an intuitive way of conceptualizing
systems that consist of autonomous actors, a major problem is the
actual design of the agent behavior. In this contribution, we examine
the potential of using agent-based learning for implementing the
agent behavior. We enhanced SeSAm, a platform for agent-based simulation,
by replacing the usual rule-based agent architecture by XCS, a well-known
learning classifier system (LCS). The resulting model is tested using
a simple evacuation scenario. The results show that on the one hand
side plausible agent behavior could be learned. On the other hand
side, though, the results are quite brittle concerning the frame
of environmental feedback, perception and action modeling.
%@ 978-3-540-87804-9
@inproceedings{Kluegl:2008:mates,
abstract = {Although multi-agent simulations are an intuitive way of conceptualizing
systems that consist of autonomous actors, a major problem is the
actual design of the agent behavior. In this contribution, we examine
the potential of using agent-based learning for implementing the
agent behavior. We enhanced SeSAm, a platform for agent-based simulation,
by replacing the usual rule-based agent architecture by XCS, a well-known
learning classifier system (LCS). The resulting model is tested using
a simple evacuation scenario. The results show that on the one hand
side plausible agent behavior could be learned. On the other hand
side, though, the results are quite brittle concerning the frame
of environmental feedback, perception and action modeling.},
added-at = {2017-03-16T11:50:55.000+0100},
address = {Kaiserslautern, Germany},
author = {Kl\"ugl, Franziska and Hatko, Reinhard and Butz, Martin V.},
biburl = {https://www.bibsonomy.org/bibtex/2dd74a8bec38f50668531268183080982/krevelen},
booktitle = {MATES 2008: Proc. 6th German Conf. on Multiagent System
Technologies},
crossref = {mates:2008},
doi = {10.1007/978-3-540-87805-6_11},
editor = {Bergmann, Ralph and Lindemann, Gabriela and Kirn, Stefan and P\v{e}chou\v{c}ek, Michal},
interhash = {df45042a25dd7fb44dc6799276c13aa3},
intrahash = {dd74a8bec38f50668531268183080982},
isbn = {978-3-540-87804-9},
keywords = {imported thesis},
month = {23-26 Sept.},
owner = {Rick},
pages = {111--122},
publisher = {Springer},
publisher_address = {Berlin},
series = {LNCS},
timestamp = {2017-03-16T11:54:14.000+0100},
title = {Agent Learning Instead of Behavior Implementation for Simulations
--- A Case Study Using Classifier Systems},
volume = 5244,
year = 2008
}