jellekok.nl

An approach to noncommunicative multiagent coordination in continuous domains

Jelle R. Kok, Matthijs T. J. Spaan, and Nikos Vlassis. An approach to noncommunicative multiagent coordination in continuous domains. In Benelearn 2002: Proceedings of the Twelfth Belgian-Dutch Conference on Machine Learning, pp. 46–52, Utrecht, The Netherlands, December 2002.

Download

pdf [126.7kB]  ps.gz [109.2kB]  

Abstract

Principled game-theoretic techniques exist for solving the problem of action coordination in a group of agents, however they typically suffer from an exponential blowup of the action space when many agents are involved. Coordination graphs (Guestrin et. al., 2002) offer tractable approximations via a context-specific decomposition into smaller coordination problems, and they are based on an iterative communication-based action selection procedure. We propose two extensions that apply when the agents are embedded in a continuous domain and/or communication is unavailable.

BibTeX Entry

@InProceedings{Kok02benelearn,
  author =       {Jelle R. Kok and Matthijs T. J. Spaan and Nikos
                  Vlassis},
  title =        {An approach to noncommunicative multiagent
                  coordination in continuous domains},
  address =      {Utrecht, The Netherlands},
  booktitle =    {Benelearn 2002: Proceedings of the Twelfth
                  Belgian-Dutch Conference on Machine Learning},
  year =         {2002},
  pages =        {46--52},
  editor =       {Marco Wiering},
  month =        dec,
  postscript =   {2002/Kok02benelearn.ps.gz},
  pdf =          {2002/Kok02benelearn.pdf},
  abstract =     {Principled game-theoretic techniques exist for
                  solving the problem of action coordination in a
                  group of agents, however they typically suffer from
                  an exponential blowup of the action space when many
                  agents are involved. Coordination graphs (Guestrin
                  et. al., 2002) offer tractable approximations via a
                  context-specific decomposition into smaller
                  coordination problems, and they are based on an
                  iterative communication-based action selection
                  procedure. We propose two extensions that apply when
                  the agents are embedded in a continuous domain
                  and/or communication is unavailable.}
}

Generated by bib2html.pl (written by Patrick Riley) on Tue Oct 31, 2006 19:33:42 UTC