Measuring Mutation Operators’ Exploration-Exploitation Behaviour and Long-Term Biases

  • James McDermott
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 8599)


We propose a simple method of directly measuring a mutation operator’s short-term exploration-exploitation behaviour, based on its transition matrix. Higher values for this measure indicate a more exploitative operator. Since operators also differ in their degree of long-term bias towards particular areas of the search space, we propose a simple method of directly measuring this bias, based on the Markov chain stationary state. We use these measures to compare numerically the behaviours of two well-known mutation operators, the genetic algorithm per-gene bitflip mutation and the genetic programming subtree mutation.


genetic programming mutation transition probability Markov chain stationary distribution operator bias 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Bertsimas, D., Tsitsiklis, J.: Simulated annealing. Statistical Science, 10–15 (1993)Google Scholar
  2. 2.
    Dignum, S., Poli, R.: Crossover, sampling, bloat and the harmful effects of size limits. In: O’Neill, M., Vanneschi, L., Gustafson, S., Esparcia Alcázar, A.I., De Falco, I., Della Cioppa, A., Tarantino, E. (eds.) EuroGP 2008. LNCS, vol. 4971, pp. 158–169. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  3. 3.
    Droste, S., Jansen, T., Wegener, I.: On the analysis of the (1+1) evolutionary algorithm. Theoretical Computer Science 276(1), 51–81 (2002)CrossRefzbMATHMathSciNetGoogle Scholar
  4. 4.
    Eiben, A.E., Aarts, E.H.L., Hee, K.M.V.: Global convergence of genetic algorithms: A Markov chain analysis. In: Schwefel, H.-P., Männer, R. (eds.) PPSN 1990. LNCS, vol. 496, pp. 3–12. Springer, Heidelberg (1991)CrossRefGoogle Scholar
  5. 5.
    Garnier, J., Kallel, L., Schoenauer, M.: Rigorous hitting times for binary mutations. Evolutionary Computation 7(2), 173–203 (1999)CrossRefGoogle Scholar
  6. 6.
    Hu, T., Payne, J.L., Banzhaf, W., Moore, J.H.: Evolutionary dynamics on multiple scales: a quantitative analysis of the interplay between genotype, phenotype, and fitness in linear genetic programming. Genetic Programming and Evolvable Machines 13(3), 305–337 (2012)CrossRefGoogle Scholar
  7. 7.
    Kemeny, J.G., Snell, J.L.: Finite Markov chains, vol. 210. Springer, New York (1976)zbMATHGoogle Scholar
  8. 8.
    Koza, J.R.: Genetic Programming: On the Programming of Computers by Means of Natural Selection. The MIT Press, Cambridge (1992)zbMATHGoogle Scholar
  9. 9.
    Levin, D.A., Peres, Y., Wilmer, E.L.: Markov chains and mixing times. AMS Bookstore (2009),
  10. 10.
    Moraglio, A., Krawiec, K., Johnson, C.G.: Geometric semantic genetic programming. In: Coello, C.A.C., Cutello, V., Deb, K., Forrest, S., Nicosia, G., Pavone, M. (eds.) PPSN 2012, Part I. LNCS, vol. 7491, pp. 21–31. Springer, Heidelberg (2012)CrossRefGoogle Scholar
  11. 11.
    Moritz, R., Ulrich, T., Thiele, L., Buerklen, S.: Mutation operator characterization: Exhaustiveness, locality, and bias. In: Congress on Evolutionary Computation. IEEE, New Orleans (2011)Google Scholar
  12. 12.
    Ni, J., Drieberg, R.H., Rockett, P.I.: The use of an analytic quotient operator in genetic programming. Transactions on Evolutionary Computation 17(1) (2013)Google Scholar
  13. 13.
    O’Reilly, U.M., Oppacher, F.: Program search with a hierarchical variable length representation: Genetic programming, simulated annealing and hill climbing. In: Davidor, Y., Schwefel, H.-P., Männer, R. (eds.) PPSN 1994. LNCS, vol. 866, pp. 397–406. Springer, Heidelberg (1994), CrossRefGoogle Scholar
  14. 14.
    Pagie, L., Hogeweg, P.: Evolutionary Consequences of Coevolving Targets. Evolutionary Computation 5, 401–418 (1997)CrossRefGoogle Scholar
  15. 15.
    Poli, R., McPhee, N.F., Rowe, J.E.: Exact schema theory and Markov chain models for genetic programming and variable-length genetic algorithms with homologous crossover. Genetic Programming and Evolvable Machines 5(1), 31–70 (2004)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2014

Authors and Affiliations

  • James McDermott
    • 1
    • 2
  1. 1.Natural Computing Research and Applications Group, Complex and Adaptive Systems LabUniversity College DublinIreland
  2. 2.Management Information Systems, Lochlann Quinn School of BusinessUniversity College DublinIreland

Personalised recommendations