Constant versus variable arity operators in genetic programming

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

  author =       "Michael McMullin and Terence Soule",
  title =        "Constant versus variable arity operators in genetic
  booktitle =    "GECCO '10: Proceedings of the 12th annual conference
                 on Genetic and evolutionary computation",
  year =         "2010",
  editor =       "Juergen Branke and Martin Pelikan and Enrique Alba and 
                 Dirk V. Arnold and Josh Bongard and 
                 Anthony Brabazon and Juergen Branke and Martin V. Butz and 
                 Jeff Clune and Myra Cohen and Kalyanmoy Deb and 
                 Andries P Engelbrecht and Natalio Krasnogor and 
                 Julian F. Miller and Michael O'Neill and Kumara Sastry and 
                 Dirk Thierens and Jano {van Hemert} and Leonardo Vanneschi and 
                 Carsten Witt",
  isbn13 =       "978-1-4503-0072-8",
  pages =        "987--988",
  keywords =     "genetic algorithms, genetic programming, Poster",
  month =        "7-11 " # jul,
  organisation = "SIGEVO",
  address =      "Portland, Oregon, USA",
  DOI =          "doi:10.1145/1830483.1830663",
  publisher =    "ACM",
  publisher_address = "New York, NY, USA",
  abstract =     "In this paper we compare typical variable arity
                 operators to constant arity operators in which
                 extraneous branches are treated as no-ops. We suggest
                 that the consistent arity implementation would perform
                 poorer than the variable arity implementation, due to a
                 large amount of non-productive changes experienced
                 during the early life of a consistent arity individual.
                 Contrary to the expected result, both algorithms
                 performed nearly identically. The consistent arity
                 population developed a better average population
                 faster, indicating that it would be a better option for
                 tasks requiring many options for success. The variable
                 arity population developed much smaller individuals on
                 average, taking up much less space. This may be partly
                 due to the large proportion of arity one operators in
                 the operator set. However, a comparison of the
                 execution times produced surprisingly mixed results,
                 with the variable arity approach sometime taking
                 significantly more time despite producing significantly
                 smaller trees.",
  notes =        "truck backing up, symbolic regression, inter-twined
                 spirals, Also known as \cite{1830663} GECCO-2010 A
                 joint meeting of the nineteenth international
                 conference on genetic algorithms (ICGA-2010) and the
                 fifteenth annual genetic programming conference

Genetic Programming entries for Michael McMullin Terence Soule