Graph Coloring with Adaptive Evolutionary Algorithms

  • Authors:
  • A. E. Eiben;J. K. Van Der Hauw;J. I. Van Hemert

  • Affiliations:
  • Leiden University;Leiden University;Leiden University

  • Venue:
  • Journal of Heuristics
  • Year:
  • 1998

Quantified Score

Hi-index 0.00

Visualization

Abstract

This paper presents the results of an experimentalinvestigation on solving graph coloring problems with EvolutionaryAlgorithms (EAs). After testing different algorithm variants weconclude that the best option is an asexual EA using order-basedrepresentation and an adaptation mechanism that periodically changesthe fitness function during the evolution. This adaptive EA isgeneral, using no domain specific knowledge, except, of course, fromthe decoder (fitness function). We compare this adaptive EA to apowerful traditional graph coloring technique DSatur and the Grouping Genetic Algorithm (GGA) on a wide range of problem instances with different size, topologyand edge density. The results show that the adaptive EA is superiorto the Grouping (GA) and outperforms DSatur on the hardest probleminstances. Furthermore, it scales up better with the problem sizethan the other two algorithms and indicates a linear computationalcomplexity.