Vidal's library
Title: Multi-Agent Patrolling with Reinforcement Learning
Author: Hugo Santana, Geber Ramalho, Vincent Corruble, and Bohdana Ratitch
Book Tittle: Proceedings of the Third International Joint Conference on Autonomous Agents and MultiAgent Systems
Pages: 1122--1129
Publisher: ACM
Year: 2004
Abstract: Patrolling tasks can be encountered in a variety of real-world domains, ranging from computer network administration and surveillance to computer wargame simulations. It is a complex multi-agent task, which usually requires agents to coordinate their decisionmaking in order to achieve optimal performance of the group as a whole. In this paper, we show how the patrolling task can be modeled as a reinforcement learning (RL) problem, allowing continuous and automatic adaptation of the agents strategies to their environment. We demonstrate that an efficient cooperative behavior can be achieved by using RL methods, such as Q-Learning, to train individual agents. The proposed approach is totally distributed, which makes it computationally efficient. The empirical evaluation proves the effectiveness of our approach, as the results obtained are substantially better than the results available so far on this domain.

Cited by 9  -  Google Scholar

@InProceedings{santana04a,
  author =	 {Hugo Santana and Geber Ramalho and Vincent Corruble
                  and Bohdana Ratitch},
  title =	 {Multi-Agent Patrolling with Reinforcement Learning},
  booktitle =	 {Proceedings of the Third International Joint
                  Conference on Autonomous Agents and MultiAgent
                  Systems},
  pages =	 {1122--1129},
  year =	 2004,
  publisher =	 {{ACM}},
  abstract =	 {Patrolling tasks can be encountered in a variety of
                  real-world domains, ranging from computer network
                  administration and surveillance to computer wargame
                  simulations. It is a complex multi-agent task, which
                  usually requires agents to coordinate their
                  decisionmaking in order to achieve optimal
                  performance of the group as a whole. In this paper,
                  we show how the patrolling task can be modeled as a
                  reinforcement learning (RL) problem, allowing
                  continuous and automatic adaptation of the agents
                  strategies to their environment. We demonstrate that
                  an efficient cooperative behavior can be achieved by
                  using RL methods, such as Q-Learning, to train
                  individual agents. The proposed approach is totally
                  distributed, which makes it computationally
                  efficient. The empirical evaluation proves the
                  effectiveness of our approach, as the results
                  obtained are substantially better than the results
                  available so far on this domain.},
  keywords =     {multiagent reinforcement learning},
  url =		 {http://jmvidal.cse.sc.edu/library/santana04a.pdf},
  comment =	 {masrg},
  googleid = 	 {n9rWT5d1qb8J:scholar.google.com/},
  cluster = 	 {13810699025048328863}
}
Last modified: Wed Mar 9 10:16:15 EST 2011