Advertisement

GA-RBF: A Self-Optimising RBF Network

  • B. Burdsall
  • C. Giraud-Carrier

Abstract

The effects of a neural network’s topology on its performance are well known, yet the question of finding optimal configurations automatically remains largely open. This paper proposes a solution to this problem for RBF networks. A self-optimising approach, driven by an evolutionary strategy, is taken. The algorithm uses output information and a computationally efficient approximation of RBF networks to optimise the K-means clustering process by co-evolving the two determinant parameters of the network’s layout: the number of centroids and the centroids’ positions. Empirical results demonstrate promise.

Keywords

Genetic Algorithm Hide Layer Radial Basis Function Network Genetic Algorithm Population Localize Receptive Field 
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]
    C. Baroglio, A. Giordiana, M. Kaiser, M. Nuttin, and R. Piola. Learning controllers for industrial robots. Machine Learning, 23(2/3):221–249, 1996.CrossRefGoogle Scholar
  2. [2]
    E. Blanzieri, A. Giordiana, and P. Katenkamp. Growing radial basis function networks. In Proceedings of the Fourth Workshop on Learning Robots, 1995.Google Scholar
  3. [3]
    B. Burdsall and C. Giraud-Carrier. Evolving fuzzy prototypes for efficient data clustering. In Proceedings of ISFL’97, 1997.Google Scholar
  4. [4]
    C. Decaestecker. Design of a neural net classifier using prototypes. Technical report, IRIDIA, Université Libre de Bruxelles, 1994.Google Scholar
  5. [5]
    L.J. Fogel. Evolutionary programming in perspective: The top down view. In J.M. Zurada et al, editors, Computational Intelligence: Imitating Life, NY, 1994. IEEE Press, Inc.Google Scholar
  6. [6]
    B. Lemarié and A-G. Debroise. A dynamical architecture for radial basis function networks. In D.W. Pearson, N.C. Steele, R.F. Albrecht, editors, Artificial Neural Networks and Genetic Algorithms, pages 305–308, Wien New York, 1995. Springer-Verlag.Google Scholar
  7. [7]
    Z. Michalewicz. Genetic Algorithms + Data Structures = Evolution Programs. Springer Verlag, 1992.Google Scholar
  8. [8]
    J. Moody and C. Darken. Learning with localized receptive fields. In Proceedings of the 1988 Connectionist Models Summer School, Pittsburgh, PA, 1988.Google Scholar
  9. [9]
    J. Moody and C. Darken. Fast learning in networks of locally-tuned processing units. Neural Computation, 1(2):281–294, 1989.CrossRefGoogle Scholar
  10. [10]
    S. Mukherjee and S. Nayar. Automatic generation of rbf networks. Technical Report CUCS-001-95, Department of Computer Science, Columbia University, 1995.Google Scholar
  11. [11]
    P.M. Murphy and D.W. Aha. UCI Repository of Machine Learning Databases. University of California, Irvine, Department of Information and Computer Science.Google Scholar
  12. [12]
    R. Neruda. Functional equivalence and genetic learning of RBF networks. In D.W. Pearson, N. C. Steele, R.F. Albrecht, editors, Artificial Neural Networks and Genetic Algorithms, pages 53–56, Wien New York, 1995. Springer-Verlag.Google Scholar
  13. [13]
    T. Poggio and F. Girosi. Networks for approximation and learning. Proceedings of IEEE, 78(9):1481–1497, 1990.Google Scholar
  14. [14]
    K. Warwick, J.D. Mason, and E.L. Sutano. Center selection for a radial basis function network. In R.F. Albrecht D.W. Pearson, N.C. Steele, editors, Artificial Neural Networks and Genetic Algorithms, pages 309–312, Wien New York, 1995. Springer-Verlag.Google Scholar

Copyright information

© Springer-Verlag Wien 1998

Authors and Affiliations

  • B. Burdsall
    • 1
  • C. Giraud-Carrier
    • 2
  1. 1.Mastère 2IA, ENST de BretagneBrest CedexFrance
  2. 2.Department of Computer ScienceUniversity of BristolBristolEngland

Personalised recommendations