Evolving Problems to Learn about Particle Swarm Optimisers and other Search Algorithms

Created by W.Langdon from gp-bibliography.bib Revision:1.4549

  author =       "W. B. Langdon and Riccardo Poli",
  title =        "Evolving Problems to Learn about Particle Swarm
                 Optimisers and other Search Algorithms",
  journal =      "IEEE Transactions on Evolutionary Computation",
  year =         "2007",
  volume =       "11",
  number =       "5",
  pages =        "561--578",
  month =        oct,
  keywords =     "genetic algorithms, genetic programming, PSO, DE,
                 CMA-ES, local search, XPS, Differential evolution (DE),
                 fitness landscapes, hill-climbers, particle swarms",
  ISSN =         "1089-778X",
  URL =          "http://www.cs.ucl.ac.uk/staff/W.Langdon/ftp/papers/langdon_2006_TEC.pdf",
  URL =          "http://ieeexplore.ieee.org/iel5/4235/26785/101109TEVC2006886448.pdf?arnumber=101109TEVC2006886448&isnumber=26785",
  DOI =          "doi:10.1109/TEVC.2006.886448",
  size =         "18 pages",
  abstract =     "We use evolutionary computation (EC) to automatically
                 find problems which demonstrate the strength and
                 weaknesses of modern search heuristics. In particular
                 we analyse Particle Swarm Optimization (PSO),
                 Differential Evolution (DE) and Covariance Matrix
                 Adaptation-Evolution Strategy (CMA-ES). Each
                 evolutionary algorithms is contrasted with the others
                 and with a robust non-stochastic gradient follower
                 (i.e. a hill climber) based on Newton-Raphson. The
                 evolved benchmark problems yield insights into the
                 operation of PSOs, illustrate benefits and drawbacks of
                 different population sizes, velocity limits and
                 constriction (friction) coefficients. The fitness
                 landscapes made by genetic programming (GP) reveal new
                 swarm phenomena, such as deception, thereby explaining
                 how they work and allowing us to devise better extended
                 particle swarm systems. The method could be applied to
                 any type of optimiser.",
  notes =        "evolving test benchmarks See also

Genetic Programming entries for William B Langdon Riccardo Poli