Advertisement

Avoiding Local Minima in ANN by Genetic Evolution

  • R. Mendes
  • P. Vale
  • J. M. Sousa
  • J. A. Roubos
Conference paper

Abstract

A novel approach is proposed to avoid the problem of local minima in neural networks learning when using the back-propagation algorithm. This problem is solved by applying a genetic evolution. The unification of both powerful technies can be viewed as an example of a mix between Darwinian and Lamarckian learning. The wine data, a high-dimensional classification problem, is given as an example.

Keywords

Artificial Neural Network Hide Layer Training Epoch Wine Data Good Classification Rate 
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]
    S. Haykin. Neural Networks — A comprehensive foundation. Macmillan College Publishing Company, 1994.Google Scholar
  2. [2]
    T. Mitchell. Machine Learning. McGraw-Hill, 1997.Google Scholar
  3. [3]
    S.-B. Cho. “Pattern recognition with neural networks combined by genetic algorithms”, Fuzzy Sets and Systems, 103:339–347, 1999.CrossRefGoogle Scholar
  4. [4]
    A. Blanco, M. Delgado, M.C. Pegalajar. “A genetic algorithm to obtain the optimal recurrent neural network”, Int. J. Approx. Reasoning, 23:67–83, 2000.CrossRefMATHGoogle Scholar
  5. [5]
    M. Setnes, J.A. Roubos. “GA-Fuzzy modeling and classification: complexity and performance”, IEEE Trans. on Fuzzy Systems, 8(5):509–522, Oct. 2000.CrossRefGoogle Scholar
  6. [6]
    UCI Repository of machine learning databases, http://www.ics.uci.edu/-mlearnl, 1998.Google Scholar

Copyright information

© Springer-Verlag Wien 2001

Authors and Affiliations

  • R. Mendes
    • 1
  • P. Vale
    • 1
  • J. M. Sousa
    • 1
  • J. A. Roubos
    • 2
  1. 1.Instituto Superior Técnico, Dept. of Mechanical Engineering/GCARTechnical University of LisbonLisbonPortugal
  2. 2.Control Laboratory, Faculty of Information, Technology and SystemsDelft University of TechnologyDelftThe Netherlands

Personalised recommendations