Competitive Environments Evolve Better Solutions for Complex Tasks

Created by W.Langdon from gp-bibliography.bib Revision:1.4420

  author =       "Peter J. Angeline and Jordan B. Pollack",
  title =        "Competitive Environments Evolve Better Solutions for
                 Complex Tasks",
  year =         "1993",
  booktitle =    "Proceedings of the 5th International Conference on
                 Genetic Algorithms, ICGA-93",
  editor =       "Stephanie Forrest",
  publisher =    "Morgan Kaufmann",
  pages =        "264--270",
  month =        "17-21 " # jul,
  address =      "University of Illinois at Urbana-Champaign",
  publisher_address = "2929 Campus Drive, Suite 260, San Mateo, CA
                 94403, USA",
  keywords =     "genetic algorithms, genetic programming",
  URL =          "",
  URL =          "",
  URL =          "",
  size =         "7 pages",
  abstract =     "In the typical genetic algorithm experiment, the
                 fitness function is constructed to be independent of
                 the contents of the population to provide a consistent
                 objective measure. Such objectivity entails significant
                 knowledge about the environment which suggests either
                 the problem has previously been solved or other
                 non-evolutionary techniques may be more efficient.
                 Furthermore, for many complex tasks an independent
                 fitness function is either impractical or impossible to
                 provide. In this paper, we demonstrate that competitive
                 fitness functions, i.e. fitness functions that are
                 dependent on the constituents of the population, can
                 provide a more robust training environment than
                 independent fitness functions. We describe three
                 differing methods for competitive fitness, and discuss
                 their respective advantages.",
  ISBN =         "1-55860-299-2",
  notes =        "very like thesis

                 One method I investigated was called competitive
                 fitness functions which is a fitness function that
                 compares performance between members of the population
                 to determine a ranking of individuals for reproduction.
                 THis obviates the need for a quantitative model of the
                 quality of solutions and replaces it with a more
                 simplistic measure of {"}x is better than y{"}. The
                 paper explores this concept using GLiB and appeared in

Genetic Programming entries for Peter John Angeline Jordan B Pollack