Solving the really hard problems with cooperative search

  • Authors:
  • Tad Hogg;Colin P. Williams

  • Affiliations:
  • Xerox Palo Alto Research Center, Palo Alto, CA;Xerox Palo Alto Research Center, Palo Alto, CA

  • Venue:
  • AAAI'93 Proceedings of the eleventh national conference on Artificial intelligence
  • Year:
  • 1993

Quantified Score

Hi-index 0.00

Visualization

Abstract

We present and experimentally evaluate the hypothesis that cooperative parallel search is well suited for hard graph coloring problems near a previously identified transition between under- and overconstrained instances. We find that simple cooperative methods can often solve such problems faster than the same number of independent agents.