Advertisement

Revisiting UCS: Description, Fitness Sharing, and Comparison with XCS

  • Albert Orriols-Puig
  • Ester Bernadó-Mansilla
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 4998)

Abstract

This paper provides a deep insight into the learning mechanisms of UCS, a learning classifier system (LCS) derived from XCS that works under a supervised learning scheme. A complete description of the system is given with the aim of being useful as an implementation guide. Besides, we review the fitness computation, based on the individual accuracy of each rule, and introduce a fitness sharing scheme to UCS. We analyze the dynamics of UCS both with fitness sharing and without fitness sharing over five binary-input problems widely used in the LCSs framework. Also XCS is included in the comparison to analyze the differences in behavior between both systems. Results show the benefits of fitness sharing in all the tested problems, specially those with class imbalances. Comparison with XCS highlights the dynamics differences between both systems.

Keywords

Minority Class Tournament Selection Class Imbalance Condition Length Optimal Population 
These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Bernadó-Mansilla, E., Garrell, J.M.: Accuracy-Based Learning Classifier Systems: Models, Analysis and Applications to Classification Tasks. Evolutionary Computation 11(3), 209–238 (2003)CrossRefGoogle Scholar
  2. 2.
    Brown, G., Kovacs, T., Marshall, J.A.R.: UCSpv: principled voting in UCS rule populations. In: GECCO 2007, pp. 1774–1781. ACM Press, New York (2007)Google Scholar
  3. 3.
    Bull, L., Hurst, J.: ZCS Redux. Evolutionary Computation 10(2), 185–205 (2002)CrossRefGoogle Scholar
  4. 4.
    Butz, M.V., Sastry, K., Goldberg, D.E.: Strong, stable, and reliable fitness pressure in XCS due to tournament selection. Genetic Programming and Evolvable Machines 6(1), 53–77 (2005)CrossRefGoogle Scholar
  5. 5.
    Butz, M.V.: Rule-Based Evolutionary Online Learning Systems: A Principled Approach to LCS Analysis and Design. In: Studies in Fuzziness and Soft Computing, vol. 109. Springer, Heidelberg (2006)Google Scholar
  6. 6.
    Butz, M.V., Goldberg, D., Tharankunnel, K.: Analysis and improvement of fitness exploration in XCS: Bounding models, tournament selection, and bilateral accuracy. Evolutionary Computation 11(3), 239–277 (2003)CrossRefGoogle Scholar
  7. 7.
    Butz, M.V., Goldberg, D.E., Lanzi, P.L.: Effect of pure error-based fitness in XCS. In: Kovacs, T., Llorà, X., Takadama, K., Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2003. LNCS (LNAI), vol. 4399, pp. 104–114. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  8. 8.
    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. 253–272. Springer, Heidelberg (2001)CrossRefGoogle Scholar
  9. 9.
    Harik, G.: Finding Multiple Solutions in Problems of Bounded Difficulty. Technical report, IlliGAL Report No. 94002, Urbana-Champaign IL 61801, USA (May 1994)Google Scholar
  10. 10.
    Japkowicz, N., Stephen, S.: The Class Imbalance Problem: A Systematic Study. Intelligent Data Analisis, 6(5), 429–450 (2002)zbMATHGoogle Scholar
  11. 11.
    Kovacs, T.: Strength or Accuracy? Fitness Calculation for Classifier Systems. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 1999. LNCS (LNAI), vol. 1813, pp. 143–160. Springer, Heidelberg (2000)CrossRefGoogle Scholar
  12. 12.
    Kovacs, T., Kerber, M.: What makes a problem hard for XCS. In: Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2000. LNCS (LNAI), vol. 1996, pp. 80–99. Springer, Heidelberg (2001)CrossRefGoogle Scholar
  13. 13.
    Kovacs, T., Kerber, M.: High Classification Accuracy does not Imply Effective Genetic Search. In: Deb, K., et al. (eds.) GECCO 2004. LNCS, vol. 3103, pp. 785–796. Springer, Heidelberg (2004)CrossRefGoogle Scholar
  14. 14.
    Lanzi, P.L.: A Study of the Generalization Capabilities of XCS. In: Bäck, T. (ed.) Proc. of the Seventh Int. Conf. on Genetic Algorithms, pp. 418–425. Morgan Kaufmann, San Francisco (1997)Google Scholar
  15. 15.
    Lanzi, P.L.: Learning Classifier Systems: A Reinforcement Learning Perspective. In: Studies in Fuzziness and Soft Computing, vol. 183, pp. 267–284. Springer, Heidelberg (2005)Google Scholar
  16. 16.
    Orriols-Puig, A., Bernadó-Mansilla, E.: The Class Imbalance Problem in UCS Classifier System: Fitness Adaptation. In: Congress on Evolutionary Computation, Edinburgh, UK, 2-5 September 2005, vol. 1, pp. 604–611. IEEE, Los Alamitos (2005)Google Scholar
  17. 17.
    Orriols-Puig, A., Bernadó-Mansilla, E.: Bounding XCS Parameters for Unbalanced Datasets. In: GECCO 2006, pp. 1561–1568. ACM Press, New York (2006)Google Scholar
  18. 18.
    Orriols-Puig, A., Bernadó-Mansilla, E.: The Class Imbalance Problem in UCS Classifier System: A Preliminary Study. In: Kovacs, T., Llorà, X., Takadama, K., Lanzi, P.L., Stolzmann, W., Wilson, S.W. (eds.) IWLCS 2003. LNCS (LNAI), vol. 4399, pp. 161–180. Springer, Heidelberg (2007)CrossRefGoogle Scholar
  19. 19.
    Quinlan, J.R.: C4.5: Programs for Machine Learning. Morgan Kaufmann Publishers, San Mateo (1995)Google Scholar
  20. 20.
    Sutton, R.S., Barto, A.G.: Reinforcement learning: An introduction. MIT Press, Cambridge (1998)Google Scholar
  21. 21.
    Vapnik, V.: The Nature of Statistical Learning Theory. Springer, New York (1995)CrossRefzbMATHGoogle Scholar
  22. 22.
    Wilson, S.W.: Classifier Fitness Based on Accuracy. Evolutionary Computation 3(2), 149–175 (1995)CrossRefGoogle Scholar
  23. 23.
    Wilson, S.W.: Generalization in the XCS Classifier System. In: Genetic Programming: Proceedings of the Third Annual Conference, pp. 665–674. Morgan Kaufmann, San Francisco (1998)Google Scholar
  24. 24.
    Wilson, S.W.: Classifiers that approximate functions. Journal of Natural Computing 1(2), 211–234 (2002)MathSciNetCrossRefzbMATHGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2008

Authors and Affiliations

  • Albert Orriols-Puig
    • 1
  • Ester Bernadó-Mansilla
    • 1
  1. 1.Grup de Recerca en Sistemes Intel·ligents Enginyeria i Arquitectura La SalleUniversitat Ramon LlullBarcelonaSpain

Personalised recommendations