Skip to main content
Log in

Two Neural Network Construction Methods

  • Published:
Neural Processing Letters Aims and scope Submit manuscript

Abstract

Two low complexity methods for neural network construction, that are applicable to various neural network models, are introduced and evaluated for high order perceptrons. The methods are based on a Boolean approximation of real-valued data. This approximation is used to construct an initial neural network topology which is subsequently trained on the original (real-valued) data. The methods are evaluated for their effectiveness in reducing the network size and increasing the network's generalization capabilities in comparison to fully connected high order perceptrons.

This is a preview of subscription content, log in via an institution to check access.

Access this article

Price excludes VAT (USA)
Tax calculation will be finalised during checkout.

Instant access to the full article PDF.

Institutional subscriptions

Similar content being viewed by others

References

  1. E. Fiesler and R. Beale (eds), “Handbook of neural computation”, ISBN: 0–7503–0312–3, Oxford University Press and IOP Publishing, 198 Madison Avenue, New York, NY 10016, 1996.

    Google Scholar 

  2. R. Reed, “Pruning algorithms - a survey”, IEEE Trans. on Neural Networks, Vol. 4, No. 5, pp. 740–747, Sep. 1993.

    Google Scholar 

  3. M. Wynne-Jones, “Constructive algorithms and pruning: improving the multi layer perceptron”, in Proceedings of the 13th IMACS World Congress on Computation and Applied Mathematics, R. Vichnevetsky and J.J.H. Miller (eds) 1991, Vol. 2, pp. 747–750, IMACS International Association for Mathematics and Computers in Simulation.

  4. T.-Y. Kwok and D.-Y. Yeung, “Constructive feedforward neural networks for regression problems: a survey”, Tech. Rep. HKUST-CS95–43, Dept. of Computer Science, University of Science and Technology, Hong Kong, 1995.

    Google Scholar 

  5. M._ Muselli, “On sequential construction of binary neural networks”, IEEE Trans. on Neural Networks, Vol. 6, No. 3, pp. 422–431, May 1995.

    Google Scholar 

  6. N.J. Redding, A. Kowalczyk and T. Downs, “Constructive higher-order network algorithm that is polynomial time”, Neural Networks, Vol. 6, No. 7, pp. 997–1010, 1993.

    Google Scholar 

  7. D.L. Gray and A.N. Michel, “A training algorithm for binary feedforward neural networks”, IEEE Trans. on Neural Networks, Vol. 3, No. 2, pp. 176–194, March 1992.

    Google Scholar 

  8. G. Fahner, N. Goerke and R. Eckmiller, “Structural adaptation of Boolean higher order neurons: classification with parsimonious topologies for superior generalization”, in Artificial Neural Networks, 2: Proceedings of the 1992 Int. Conference on Artificial Neural Networks, I. Alexander and J. Taylor, Eds., Amsterdam, The Netherlands, 1992, Vol. 1, pp. 285–288, Elsevier.

    Google Scholar 

  9. Y. Crama, P.L. Hammer and T. Ibaraki, “Cause-effect relationships and partially defined boolean functions”, Rutcor research report # 39- 88, Rutcor, Rutgers University, New Brunswick, N.J. 08903, USA, Aug. 1988.

    Google Scholar 

  10. D.E. Rumelhart, J.L. McClelland and the PDP Research Group, Parallel Distributed Processing: Explorations in the Microstructure of Cognition, Vol. 1: Foundations, The MIT Press, Cambridge, Massachusetts, 1986.

    Google Scholar 

  11. P.M. Murphy and D.W. Aha (librarians), UCI Repository of Machine Learning Databases, UCI Repository, ftp access ftp.ics.uci.edu: pub/machine-learning-databases, University of California, Dept. of Information and Computer Science, 1994.

  12. G. Thimmand and E. Fiesler, “Evaluating pruning methods”, in International Symposium on Artificial Neural Networks, National Chiao-Tung University, Hsinchu, Taiwan, ROC, Dec. 1995, pp. A220–25.

Download references

Author information

Authors and Affiliations

Authors

Rights and permissions

Reprints and permissions

About this article

Cite this article

Thimm, G., Fiesler, E. Two Neural Network Construction Methods. Neural Processing Letters 6, 25–31 (1997). https://doi.org/10.1023/A:1009632505828

Download citation

  • Issue Date:

  • DOI: https://doi.org/10.1023/A:1009632505828

Navigation