Coevolution of Fitness Predictors

Created by W.Langdon from gp-bibliography.bib Revision:1.4215

  title =        "Coevolution of Fitness Predictors",
  author =       "Michael D. Schmidt and Hod Lipson",
  journal =      "IEEE Transactions on Evolutionary Computation",
  year =         "2008",
  month =        dec,
  volume =       "12",
  number =       "6",
  pages =        "736--749",
  keywords =     "genetic algorithms, genetic programming, approximation
                 theory, evolutionary computation, regression analysis
                 accuracy loss, evolutionary progress, fitness
                 evaluation cost reduction, fitness predictors, model
                 approximation level, model learning investment,
                 solution bloat reduction, symbolic regression problem",
  ISSN =         "1089-778X",
  DOI =          "doi:10.1109/TEVC.2008.919006",
  abstract =     "We present an algorithm that coevolves fitness
                 predictors, optimized for the solution population,
                 which reduce fitness evaluation cost and frequency,
                 while maintaining evolutionary progress. Fitness
                 predictors differ from fitness models in that they may
                 or may not represent the objective fitness, opening
                 opportunities to adapt selection pressures and
                 diversify solutions. The use of coevolution addresses
                 three fundamental challenges faced in past fitness
                 approximation research: 1) the model learning
                 investment; 2) the level of approximation of the model;
                 and 3) the loss of accuracy. We discuss applications of
                 this approach and demonstrate its impact on the
                 symbolic regression problem. We show that coevolved
                 predictors scale favorably with problem complexity on a
                 series of randomly generated test problems. Finally, we
                 present additional empirical results that demonstrate
                 that fitness prediction can also reduce solution bloat
                 and find solutions more reliably.",
  notes =        "Also known as \cite{4476145} Three populations, pop
                 sizes = 128, 8, 10. Different representation and GA in
                 each. Comparsion with SAWs, GGGP, TAGP3
                 \cite{hoai:2002:stsrpwtgggptcr}, etc etc


Genetic Programming entries for Michael D Schmidt Hod Lipson