Enhanced Radial Basis Function Neural Network Design Using Parallel Evolutionary Algorithms

  • Elisabet Parras-Gutierrez
  • Maribel Isabel Garcia-Arenas
  • Victor M. Rivas-Santos
Part of the Communications in Computer and Information Science book series (CCIS, volume 43)


In this work SymbPar, a parallel co-evolutionary algorithm for automatically design the Radial Basis Function Networks, is proposed. It tries to solve the problem of huge execution time of Symbiotic_CHC_RBF, in which method are based. Thus, the main goal of SymbPar is to automatically design RBF neural networks reducing the computation cost and keeping good results with respect to the percentage of classification and net size. This new algorithm parallelizes the evaluation of the individuals using independent agents for every individual who should be evaluated, allowing to approach in a future bigger size problems reducing significantly the necessary time to obtain the results. SymbPar yields good results regarding the percentage of correct classified patterns and the size of nets, reducing drastically the execution time.


Neural networks evolutionary algorithms parallelization co-evolution 


Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.


  1. 1.
    Arenas, M.G., Dolin, B., Merelo, J.J., Castillo, P.A., Fernandez, I., Schoenauer, M.: JEO: Java evolving objects. In: GECCO2: Proceedings of the Genetic and Evolutionary Computation Conference (2002)Google Scholar
  2. 2.
    Bethke, A.D.: Comparison of genetic algorithms and gradient-based optimizers on parallel processors: Efficiency of use of processing capacity. Tech. Rep., University of Michigan, Ann Arbor, Logic of Computers Group (1976)Google Scholar
  3. 3.
    Castillo, P.A., et al.: G-Prop: Global optimization of multilayer perceptrons using GAs. Neurocomputing 35, 149–163 (2000)CrossRefzbMATHGoogle Scholar
  4. 4.
    Eshelman, L.J.: The CHC adptive search algorithm: How to have safe search when engaging in nontraditional genetic recombination. In: First Workshop on Foundations of Genetic Algorithms, pp. 265–283. Morgan Kaufmann, San Francisco (1991)Google Scholar
  5. 5.
    Harpham, C., et al.: A review of genetic algorithms applied to training radial basis function networks. Neural Computing & Applications 13, 193–201 (2004)CrossRefGoogle Scholar
  6. 6.
    Jelasity, M., Preub, M., Paechter, B.: A scalable and robust framework for distributed application. In: Proc. on Evolutionary Computation, pp. 1540–1545 (2002)Google Scholar
  7. 7.
    Kriegel, H., Borgwardt, K., Kroger, P., Pryakhin, A., Schubert, M., Zimek, A.: Future trends in data mining. Data Mining and Knowledge Discovery: An International Journal 15(1), 87–97 (2007)MathSciNetCrossRefGoogle Scholar
  8. 8.
    Mayer, A.H.: Symbiotic Coevolution of Artificial Neural Networks and Training Data Sets. LNCS, pp. 511–520. Springer, Heidelberg (1998)Google Scholar
  9. 9.
    Merelo, J., Prieto, A.: G-LVQ, a combination of genetic algorithms and LVQ. In: Artificial Neural Nets and Genetic Algorithms, pp. 92–95. Springer, Heidelberg (1995)CrossRefGoogle Scholar
  10. 10.
    Paredis, J.: Coevolutionary Computation. Artificial Life, 355–375 (1995)Google Scholar
  11. 11.
    Parras-Gutierrez, E., Rivas, V.M., Merelo, J.J., del Jesus, M.J.: Parameters estimation for Radial Basis Function Neural Network design by means of two Symbiotic algorithms. In: ADVCOMP 2008, pp. 164–169. IEEE computer society, Los Alamitos (2008)Google Scholar
  12. 12.
    Parras-Gutierrez, E., Rivas, V.M., Merelo, J.J., del Jesus, M.J.: A Symbiotic CHC Co-evolutionary algorithm for automatic RBF neural networks design. In: DCAI 2008, Advances in Softcomputing, Salamanca, pp. 663–671 (2008) ISSN: 1615-3871Google Scholar
  13. 13.
    Mitchell Potter, A., De Jong, K.A.: Evolving Neural Networkds with Collaborative Species. In: Proc. of the Computer Simulation Conference (1995)Google Scholar
  14. 14.
    Rivas, V.M., Merelo, J.J., Castillo, P.A., Arenas, M.G., Castellanos, J.G.: Evolving RBF neural networks for time-series forecasting with EvRBF. Information Sciences 165(3-4), 207–220 (2004)MathSciNetCrossRefGoogle Scholar
  15. 15.
    Rivas, V.M., Garcia-Arenas, I., Merelo, J.J., Prieto, A.: EvRBF: Evolving RBF Neural Networks for Classification Problems. In: Proceedings of the International Conference on Applied Informatics and Communications, pp. 100–106 (2007)Google Scholar
  16. 16.
    Rivera Rivas, A.J., Rojas Ruiz, I., Ortega Lopera, J., del Jesus, M.J.: Co-evolutionary Algorithm for RBF by Self-Organizing Population of Neurons. In: Mira, J., Álvarez, J.R. (eds.) IWANN 2003. LNCS, vol. 2686, pp. 470–477. Springer, Heidelberg (2003)CrossRefGoogle Scholar
  17. 17.
    Ros, F., Pintore, M., Deman, A., Chrtien, J.R.: Automatical initialization of RBF neural networks. In: Chemometrics and intelligent laboratory systems, vol. 87, pp. 26–32. Elsevier, Amsterdam (2007)Google Scholar
  18. 18.
    Schwaiger, R., Mayer, H.A.: Genetic algorithms to create training data sets for artificial neural networks. In: Proc. of the 3NWGA, Helsinki, Finland (1997)Google Scholar
  19. 19.
    Thompson, J.N.: The Geographic Mosaic of Coevolution. University of Chicago Press, Chicago (2005)Google Scholar
  20. 20.
    Tomassini, M.: Parallel and distributed evolutionary algorithms: A review. In: Miettinen, K., et al. (eds.) Evolutionary Algorithms in Engineering and Computer Science, pp. 113–133. J. Wiley and Sons, Chichester (1999)Google Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2009

Authors and Affiliations

  • Elisabet Parras-Gutierrez
    • 1
  • Maribel Isabel Garcia-Arenas
    • 2
  • Victor M. Rivas-Santos
    • 1
  1. 1.Department of Computer SciencesJaenSpain
  2. 2.Department of Computer Architecture and TechnologyGranadaSpain

Personalised recommendations