Advertisement

Adapting to Complexity During Search in Combinatorial Landscapes

  • Taras P. Riopka
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2611)

Abstract

Fitness landscape complexity in the context of evolutionary algorithms can be considered to be a relative term due to the complex interaction between search strategy, problem difficulty and problem representation. A new paradigm for genetic search referred to as the Collective Learning Genetic Algorithm (CLGA) has been demonstrated for combinatorial optimization problems which utilizes genotypic learning to do recombination based on a cooperative exchange of knowledge (instead of symbols) between interacting chromosomes. There is evidence to suggest that the CLGA is able to modify its recombinative behavior based on the consistency of the information in its environment, specifically, the observed fitness landscape. By analyzing the structure of the evolving individuals, a landscape-complexity measure is extracted a posteriori and then plotted for various types of example problems. This paper presents preliminary results that show that the CLGA appears to adapt its search strategy to the fitness landscape induced by the CLGA itself, and hence relative to the landscape being searched.

Keywords

Feature Detector Problem Representation Fitness Landscape Chromosome Site Problem Difficulty 
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.
    Davidor, Y. (1991). Epistasis Variance. In Foundations of Genetic Algorithms Morgan Kaufmann.Google Scholar
  2. 2.
    Heckendorn, R.B., Whitley, D. (1999). Walsh Functions and Predicting Problem Complexity. Evol. Comp., Vol. 7, No. 1, pp. 69–101. MIT Press.CrossRefGoogle Scholar
  3. 3.
    Jones, T., Forrest, S. (1995). Fitness Distance Correlation As a Measure of Problem Difficulty for Genetic Algorithms. In Proc. 6th Int. Conf. on GAs, pp. 184–192. Morgan Kaufmann.Google Scholar
  4. 4.
    Rose, H., Ebeling, W., Asselmeyer, T. (1996). The Density of States-a Measure of the Difficulty of Optimization Problems. In PPSN-IV, Springer-Verlag.Google Scholar
  5. 5.
    Reeves, C. R. (1999). Predictive Measures for Problem Difficulty. In Proc. 1999 Congress on Evol. Comp., pp. 736–743. IEEE Press.Google Scholar
  6. 6.
    Back, T. (1997). Self Adaptation. In The Handbook of Evolutionary Computation, pp. 1–23. IOP Publishing and Oxford University Press.Google Scholar
  7. 7.
    Goldberg, Korb, D.E., and Deb, K., (1989). Messy Genetic Algorithms: Motivation, Analysis, and First Results. Complex Systems, Vol. 3, pp. 493–530.zbMATHMathSciNetGoogle Scholar
  8. 8.
    Hinterding, R., Michalewicz, Z., Eiben, A.E. (1997). Adaptation in Evolutionary Computation: A Survey. In Proc. 4th Int. Conf. on Evol. Comp. pp. 65–69.Google Scholar
  9. 9.
    Riopka, T. P. (2002). Intelligent Recombination Using Genotypic Learning in a Collective Learning Genetic Algorithm, Doctoral dissertation, GWU, Washington, DC.Google Scholar
  10. 10.
    Riopka, T.P., Bock, P. (2000). Intelligent Recombination Using Individual Learning in a CLGA, In Proc. Genetic and Evol. Comp. Conf., pp. 104–111. Morgan Kaufmann.Google Scholar
  11. 11.
    Bock, P. (1993). The Emergence of Artificial Cognition. World Sci. Pub. Co.Google Scholar
  12. 12.
    De Jong, K.A. (1993). Genetic Algorithms are NOT Function Optimizers. In Foundations of Genetic Algorithms 2. Morgan Kaufmann.Google Scholar
  13. 13.
    DeJong, K.A., Potter, M.A., Spears, W.M. (1997). Using Problem Generators to Explore the Effects of Epistasis. In Proc. 7th Int. Conf. on Genetic Algorithms. Morgan Kaufmann.Google Scholar
  14. 14.
    Heckendorn, R.B., Rana, S. and Whitley, L.D. (1998). Test Function Generators as Embedded Landscapes. In Foundations of Genetic Algorithms 5, pp. 183–198. Morgan Kaufmann.Google Scholar
  15. 15.
    Crawford, J.A., Auton, L.D. (1996). Experimental Results on the Crossover Point in Random 3SAT. Art. Int. Vol. 81, No. 31.Google Scholar
  16. 16.
    Gomes, C.P., Selman, B. (2002). Satisfied with Physics. Science. Vol. 297 No. 5582.Google Scholar
  17. 17.
    Stephens, C.R. (1999). “Effective” Fitness Landscapes for Evolutionary Systems. In Proc. 1999 Congress on Evol. Comp., pp. 703–714. IEEE Press.Google Scholar
  18. 18.
    Wolpert, D.H., Macready, W.G. (1997). No Free Lunch Theorems for Optimization. IEEE Trans. Evol. Comp., vol. 1, no. 1, pp. 67–82.CrossRefGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2003

Authors and Affiliations

  • Taras P. Riopka
    • 1
  1. 1.Department of Computer Science and Engineering 200 Packard LabLehigh UniversityBethlehem

Personalised recommendations