Advertisement

Evolution Strategy with Neighborhood Attraction Using a Neural Gas Approach

  • Jutta Huhse
  • Thomas Villmann
  • Peter Merz
  • Andreas Zell
Conference paper
Part of the Lecture Notes in Computer Science book series (LNCS, volume 2439)

Abstract

In evolution strategies with neighborhood attraction (EN) the concepts of neighborhood cooperativeness and learning rules known from neural maps are transferred onto the individuals of evolution strategies. A previous approach, which utilized a neighborhood relationship adapted from self-organizing maps (SOM), appeared to perform as well as or even better than comparable conventional evolution strategies on a variety of common test functions. In this contribution, an EN with a new neighborhood relationship and learning rule based on the idea of neural gas is introduced. Its performance is compared to the SOM-like approach, using the same test functions. It is shown that the neural gas approach is considerably faster in finding the optimum than the SOM approach, although the latter seems to be more robust for multi-modal problems.

Keywords

Evolution Strategy Learning Rule Neighborhood Relationship Neighborhood Attraction Fuzzy Pattern Recognition 
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.
    Pal, S. K., Mitra, S.: Neuro fuzzy pattern recognition: methods in soft computing. Wiley, New York (1999)Google Scholar
  2. 2.
    Kohonen, T.: Self-Organizing Maps. Springer Verlag, Berlin (1995)Google Scholar
  3. 3.
    Martinetz, T., Schulten, K.: Topology representing networks. Neural Networks 7 (1994)Google Scholar
  4. 4.
    Huhse, J., Zell, A.: Evolution strategy with neighborhood attraction. In Bothe, H., Rojas, R., eds.: Proceedings of the Second ICSC Symposium on Neural Computation—NC 2000, ICSC Academic Press, Canada/Switzerland (2000) 363–369Google Scholar
  5. 5.
    Martinetz, T., Schulten, K.: A “Neural-Gas” network learns topologies. In Kohonen, T., Mäkisara, K., Simula, O., Kangas, J., eds.: Proc. International Conference on Artificial Neural Networks (Espoo, Finland). Volume I., Amsterdam, Netherlands, North-Holland (1991) 397–402Google Scholar
  6. 6.
    Martinetz, T. M., Berkovich, S. G., Schulten, K. J.: ‘Neural-gas’ network for vector quantization and its application to time-series prediction. IEEE Trans. on Neural Networks 4 (1993) 558–569CrossRefGoogle Scholar
  7. 7.
    Villmann, T.: Evolutionary algorithms and neural networks in hybrid systems. In Verleysen, M., ed.: Proceedings of 9th European Symposium on Artificial Neural Networks—ESANN’2001, Evere, Belgium, D-facto (2001)Google Scholar
  8. 8.
    Huhse, J., Zell, A.: Investigating the influence of the neighborhood attraction factor to evolution strategies with neighborhood attraction. In Verleysen, M., ed.: Proceedings of 9th European Symposium on Artificial Neural Networks—ESANN’2001, Evere, Belgium, D-facto (2001) 179–184Google Scholar
  9. 9.
    Schwefel, H. P.: Numerische Optimierung von Computer-Modellen mittels der Evolutionsstrategie. Birkhäuser, Basel, Stuttgart (1977) Volume 26 of Interdisciplinary Systems Research, German.Google Scholar
  10. 10.
    Altenberg, L.: The evolution of evolvability in genetic programming. In Kinnear, K. E., ed.: Advances in Genetic Programming. Complex Adaptive Systems, Cambridge, MIT Press (1994) 47–74Google Scholar
  11. 11.
    deJong, K.: An analysis of the behaviour of a class of genetic adaptive systems. Master’s thesis, University of Michigan (1975)Google Scholar
  12. 12.
    Bäck, T.: A user’s guide to genesys 1.0. Technical report, University of Dortmund, Department of Computer ScienceGoogle Scholar
  13. 13.
    Schwefel, H. P.: Evolution and Optimum Seeking. John Wiley and Sons, New York (1995)Google Scholar
  14. 14.
    Schwefel, H. P.: Numerical Optimization of Computer Models. Wiley, Chichester (1981)zbMATHGoogle Scholar
  15. 15.
    Ackley, D. H.: A connectionist machine for genetic hillclimbing. Kluwer Academic Publishers, Boston (1987)Google Scholar
  16. 16.
    Schwefel, H. P.: Evolutionary learning optimum-seeking on parallel computer architectures. In Sydow, A., Tzafestas, S. G., Vichnevetsky, R., eds.: Proceedings of the International Symposium on Systems Analysis and Simulation 1988, I: Theory and Foundations, Akademie der Wissenschaften der DDR, Akademie-Verlag, Berlin (1988) 217–225Google Scholar
  17. 17.
    Fletcher, R., Powell, M. J. D.: A rapidly convergent descent method for minimization. Comp. J. 6 (1963) 163–168zbMATHMathSciNetGoogle Scholar
  18. 18.
    Törn, A., Žilinskas, A.: Global Optimization. Volume 350 of Lecture Notes in Computer Science. Springer, Berlin (1989)zbMATHGoogle Scholar
  19. 19.
    Galar, R.: Simulation of local evolutionary dynamics of small populations. Biological Cybernetics 65 (1991) 37–45CrossRefGoogle Scholar
  20. 20.
    Schwefel, H. P.: Numerische Optimierung von Computer-Modellen mittels der Evolutionsstrategie. Volume 26 of Interdisciplinary systems research. Birkhäuser, Basel (1977)Google Scholar
  21. 21.
    Hansen, N., Ostermeier, A., Gawelczyk, A.: Über die Adaptation von allgemeinen, Koordinatensystem-unabhängigen, normalverteilten Mutationen in der Evolutionsstrategie: Die Erzeugendensystemadaption. Technical report, Technische Universität Berlin (1995)Google Scholar
  22. 22.
    Ostermeier, A., Gawelcyk, A., Hansen, N.: A derandomized approach to selfadaptation of evolution strategies. Evolutionary Computation 2 (1995) 369–380CrossRefGoogle Scholar
  23. 23.
    Huhse, J., Zell, A.: Evolution strategy with neighborhood attraction—A robust evolution strategy. In Spector, L., Goodman, E. D., Wu, A., Langdon, W. B., Voigt, H. M., Gen, M., Sen, S., Dorigo, M., Pezeshk, S., Garzon, M. H., Burke, E., eds.: Proceedings of the Genetic and Evolutionary Computation Conference (GECCO-2001), San Francisco, California, USA, Morgan Kaufmann (2001) 1026–1033Google Scholar
  24. 24.
    Huhse, J., Villmann, T., Zell, A.: Investigation of the neighborhood attraction evolutionary algorithm based on neural gas. In: Proceedings of the Sixth International Conference on Neural Networks and Soft Computing (ICNNSC) (to appear). Lecture Notes in Computer Science, Springer (2002)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2002

Authors and Affiliations

  • Jutta Huhse
    • 1
  • Thomas Villmann
    • 2
  • Peter Merz
    • 1
  • Andreas Zell
    • 1
  1. 1.Inst. of Computer ScienceUniversity of TübingenTübingenGermany
  2. 2.Clinic for Psychotherapy and Psychosomatic MedicineUniversity of LeipzigLeipzigGermany

Personalised recommendations