@InProceedings{makar01a,
  author =	 {Rajbala Makar and Sridhar Mahadevan and Mohammad
                  Ghavamzadeh},
  title =	 {Hierarchical MultiAgent Reinforcement Learning},
  booktitle =	 {Proceedings of the Fifth International Conference on
                  Autonomous Agents},
  pages =	 {246--253},
  year =	 2001,
  abstract =	 {In this paper we investigate the use of hierarchical
                  rein- forcement learning to speed up the acquisition
                  of cooper- ative multi-agent tasks. We extend the
                  MAXQ framework to the multi-agent case. Each agent
                  uses the same MAXQ hierarchy to decompose a task
                  into sub-tasks. Learning is decentralized, with each
                  agent learning three interrelated skills: how to
                  perform subtasks, which order to do them in, and how
                  to coordinate with other agents. Coordination skills
                  among agents are learned by using joint actions at
                  the highest level(s) of the hierarchy. The Q nodes
                  at the high- est level(s) of the hierarchy are
                  congured to represent the joint task-action space
                  among multiple agents. In this ap- proach, each
                  agent only knows what other agents are doing at the
                  level of sub-tasks, and is unaware of lower level
                  (prim- itive) actions. This hierarchical approach
                  allows agents to learn coordination faster by
                  sharing information at the level of sub-tasks,
                  rather than attempting to learn coordination taking
                  into account primitive joint state-action values. We
                  apply this hierarchical multi-agent reinforcement
                  learning algorithm to a complex AGV scheduling task
                  and compare its performance and speed with other
                  learning approaches, including at multi-agent,
                  single agent using MAXQ, selsh multiple agents using
                  MAXQ (where each agent acts inde- pendently without
                  communicating with the other agents), as well as
                  several well-known AGV heuristics like "rst come rst
                  serve", "highest queue rst" and "nearest station
                  rst". We also compare the tradeos in learning speed
                  vs. perfor- mance of modeling joint action values at
                  multiple levels in the MAXQ hierarchy.},
  keywords =     {multiagent reinforcement learning},
  url = 	 {http://jmvidal.cse.sc.edu/library/makar01a.pdf},
  cluster = 	 {2242243449876899562}
}