How Fitness Estimates Interact with Reproduction Rates: Towards Variable Offspring Set Sizes in XCSF

  • Patrick O. Stalph
  • Martin V. Butz
Part of the Lecture Notes in Computer Science book series (LNCS, volume 6471)


Despite many successful applications of the XCS classifier system, a rather crucial aspect of XCS’ learning mechanism has hardly ever been modified: exactly two classifiers are reproduced when XCSF’s iterative evolutionary algorithm is applied in a sampled problem niche. In this paper, we investigate the effect of modifying the number of reproduced classifiers. In the investigated problems, increasing the number of reproduced classifiers increases the initial learning speed. In less challenging approximation problems, also the final approximation accuracy is not affected. In harder problems, however, learning may stall, yielding worse final accuracies. In this case, over-reproductions of inaccurate, ill-estimated, over-general classifiers occur. Since the quality of the fitness signal decreases if there is less time for evaluation, a higher reproduction rate can deteriorate the fitness signal, thus—dependent on the difficulty of the approximation problem—preventing further learning improvements. In order to speed-up learning where possible while still assuring learning success, we propose an adaptive offspring set size that may depend on the current reliability of classifier parameter estimates. Initial experiments with a simple offspring set size adaptation show promising results.


LCS XCS Reproduction Selection Pressure 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Holland, J.H.: Adaptation. In: Progress in Theoretical Biology, vol. 4, pp. 263–293. Academic Press, New York (1976)Google Scholar
  2. 2.
    Holland, J.H.: Properties of the bucket brigade algorithm. In: Proceedings of the 1st International Conference on Genetic Algorithms, Hillsdale, NJ, USA, pp. 1–7. L. Erlbaum Associates Inc., Mahwah (1985)Google Scholar
  3. 3.
    Widrow, B., Hoff, M.E.: Adaptive switching circuits. Western Electronic Show and Convention, Convention Record, Part 4, 96–104 (1960)Google Scholar
  4. 4.
    Wilson, S.W.: Classifier fitness based on accuracy. Evolutionary Computation 3(2), 149–175 (1995)CrossRefGoogle Scholar
  5. 5.
    Lanzi, P.L., Loiacono, D., Wilson, S.W., Goldberg, D.E.: Prediction update algorithms for XCSF: RLS, Kalman filter, and gain adaptation. In: GECCO 2006: Proceedings of the 8th Annual Conference on Genetic and Evolutionary Computation, pp. 1505–1512. ACM, New York (2006)Google Scholar
  6. 6.
    Drugowitsch, J., Barry, A.: A formal framework and extensions for function approximation in learning classifier systems. Machine Learning 70, 45–88 (2008)CrossRefGoogle Scholar
  7. 7.
    Orriols-Puig, A., Bernadó-Mansilla, E.: Bounding XCS’s parameters for unbalanced datasets. In: GECCO 2006: Proceedings of the 8th Annual Conference on Genetic and Evolutionary Computation, pp. 1561–1568. ACM, New York (2006)Google Scholar
  8. 8.
    Wilson, S.W.: Get real! XCS with continuous-valued inputs. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 1999. LNCS (LNAI), vol. 1813, pp. 209–219. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  9. 9.
    Wilson, S.W.: Classifiers that approximate functions. Natural Computing 1, 211–234 (2002)MathSciNetCrossRefzbMATHGoogle Scholar
  10. 10.
    Butz, M.V., Lanzi, P.L., Wilson, S.W.: Function approximation with XCS: Hyperellipsoidal conditions, recursive least squares, and compaction. IEEE Transactions on Evolutionary Computation 12, 355–376 (2008)CrossRefGoogle Scholar
  11. 11.
    Butz, M.V., Wilson, S.W.: An algorithmic description of XCS. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2000. LNCS (LNAI), vol. 1996, pp. 267–274. Springer, Heidelberg (2001)Google Scholar
  12. 12.
    Butz, M.V., Kovacs, T., Lanzi, P.L., Wilson, S.W.: Toward a theory of generalization and learning in XCS. IEEE Transactions on Evolutionary Computation 8, 28–46 (2004)CrossRefGoogle Scholar
  13. 13.
    Butz, M.V., Goldberg, D.E., Tharakunnel, K.: Analysis and improvement of fitness exploitation in XCS: Bounding models, tournament selection, and bilateral accuracy. Evolutionary Computation 11, 239–277 (2003)CrossRefGoogle Scholar
  14. 14.
    Butz, M.V., Goldberg, D.E., Lanzi, P.L., Sastry, K.: Problem solution sustenance in XCS: Markov chain analysis of niche support distributions and the impact on computational complexity. Genetic Programming and Evolvable Machines 8, 5–37 (2007)CrossRefGoogle Scholar
  15. 15.
    Butz, M.V., Goldberg, D.E., Lanzi, P.L.: Bounding learning time in XCS. In: Deb, K., et al. (eds.) GECCO 2004. LNCS, vol. 3103, pp. 739–750. Springer, Heidelberg (2004)CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2010

Authors and Affiliations

  • Patrick O. Stalph
    • 1
  • Martin V. Butz
    • 1
  1. 1.Department of Cognitive Psychology IIIUniversity of WürzburgWürzburgGermany

Personalised recommendations