Advertisement

Vector Perceptron Learning Algorithm Using Linear Programming

  • Vladimir Kryzhanovskiy
  • Irina Zhelavskaya
  • Anatoliy Fonarev
Part of the Lecture Notes in Computer Science book series (LNCS, volume 7553)

Abstract

Application of Linear Programming for binary perceptron learning allows reaching theoretical maximum loading of the perceptron that had been predicted by E. Gardner. In the present paper the idea of learning using Linear Programming is extended to vector multistate neural networks. Computer modeling shows that the probability of false identification for the proposed learning rule decreases by up to 50 times compared to the Hebb one.

Keywords

vector neural networks simplex-method linear programming 

Preview

Unable to display preview. Download preview PDF.

Unable to display preview. Download preview PDF.

References

  1. 1.
    Kanter, I.: Potts-glass models of neural networks. Physical Review A 37(7), 2739–2742 (1988)CrossRefMathSciNetGoogle Scholar
  2. 2.
    Cook, J.: The mean-field theory of a Q-state neural network model. Journal of Physics A 22, 2000–2012 (1989)CrossRefGoogle Scholar
  3. 3.
    Bolle, D., Dupont, P., Huyghebaert, J.: Thermodynamics properties of the q-state Potts-glass neural network. Phys. Rew. A 45, 4194–4197 (1992)CrossRefGoogle Scholar
  4. 4.
    Wu, F.: The Potts model. Review of Modern Physics 54, 235–268 (1982)CrossRefGoogle Scholar
  5. 5.
    Kryzhanovsky, B., Mikaelyan, A.: On the Recognition Ability of a Neural Network on Neurons with Parametric Transformation of Frequencies. Doklady Mathematics 65(2), 286–288 (2002)Google Scholar
  6. 6.
    Kryzhanovsky, B., Kryzhanovsky, V., Litinskii, L.: Machine Learning in Vector Models of Neural Networks. In: Koronacki, J., Raś, Z.W., Wierzchoń, S.T., Kacprzyk, J. (eds.) Advances in Machine Learning II. SCI, vol. 263, pp. 427–443. Springer, Heidelberg (2010)CrossRefGoogle Scholar
  7. 7.
    Kryzhanovskiy, V.: Binary Patterns Identification by Vector Neural Network with Measure of Proximity between Neuron States. In: Honkela, T. (ed.) ICANN 2011, Part II. LNCS, vol. 6792, pp. 119–126. Springer, Heidelberg (2011)Google Scholar
  8. 8.
    Austin, J., Turner, A., Lees, K.: Chemical Structure Matching Using Correlation Matrix Memories. In: International Conference on Artificial Neural Networks, IEE Conference Publication 470, Edinburgh, UK, September 7-10. IEE, London (1999)Google Scholar
  9. 9.
    Krauth, W., Mezard, M.: Learning algorithms with optimal stability in neural networks. J. Phys. A: Math. Gen. 20, L745–L752 (1987)Google Scholar
  10. 10.
    Kryzhanovskiy, V., Zhelavskaya, I., Karandashev, J.: Binary Perceptron Learning Algorithm Using Simplex-Method. In: Rutkowski, L., Korytkowski, M., Scherer, R., Tadeusiewicz, R., Zadeh, L.A., Zurada, J.M. (eds.) ICAISC 2012, Part I. LNCS, vol. 7267, pp. 111–118. Springer, Heidelberg (2012)CrossRefGoogle Scholar
  11. 11.
    Kryzhanovsky, B., Kryzhanovsky, V.: Binary Optimization: On the Probability of a Local Minimum Detection in Random Search. In: Rutkowski, L., Tadeusiewicz, R., Zadeh, L.A., Zurada, J.M. (eds.) ICAISC 2008. LNCS (LNAI), vol. 5097, pp. 89–100. Springer, Heidelberg (2008)CrossRefGoogle Scholar
  12. 12.
    Center of Optical-Neural Technologies, http://www.niisi.ru/iont/downloads/km/
  13. 13.
    Gardner, E., Derrida, B.: Optimal storage properties of neural network models. J. Phys. A: Math. Gen. 21, 271–284 (1988)CrossRefMathSciNetGoogle Scholar

Copyright information

© Springer-Verlag Berlin Heidelberg 2012

Authors and Affiliations

  • Vladimir Kryzhanovskiy
    • 1
  • Irina Zhelavskaya
    • 1
  • Anatoliy Fonarev
    • 2
  1. 1.Scientific Research Institute for System AnalysisRussian Academy of SciencesMoscowRussia
  2. 2.Department of Engineering and ScienceCUNY City University of New YorkUSA

Personalised recommendations