Skip to main content
SpringerLink
Log in

Asymptotic performances of a constructive algorithm

  • Published:
Neural Processing Letters Aims and scope Submit manuscript

Abstract

We present a numerical study of a neural tree learning algorithm, the “triolearning≓ strategy. We study the behaviour of the algorithm as a function of the size of the training set. The results show that a limited number of examples can be used to estimate both the network performance and the network complexity that would result from running the algorithm on a large data set.

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

Access this article

Log in via an institution

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. J. A. Sirat, J.-P. Nadal. Neural trees: a new tool for classification,Network, vol. 1, pp. 423–438, 1990.

    MathSciNet  Google Scholar 

  2. M. Golea, M. Marchand. A growth algorithm for neural network decision trees,Europhys. Lett., vol. 12, pp. 105–109, 1990.

    Google Scholar 

  3. M. Frean. The Upstart algorithm: a method for constructing and training feedforward neural networks,Neural Computation, vol. 2, pp. 198–209, 1990.

    Google Scholar 

  4. L. Breiman et al.Classification and regression trees. Wadsworth International Group, Belmont, CA, 1984.

    Google Scholar 

  5. J.R. Quinlan. Induction of Decision Trees,Machine Learning, vol. 1, pp. 81–106, 1986.

    Google Scholar 

  6. C. Cortès et al. Learning curves: asymptotic values and rates of convergence. In:Neural Networks for Computing Conference, Snowbird, Utah, 1993.

  7. V. Vapnik, E. Levin, Y. Le Cun. Measuring the VC-Dimension of a learning machine,Neural Computation, vol. 6, pp. 851–876, 1994.

    Google Scholar 

  8. F. d'Alché-Buc, D. Zwierski, J.-P. Nadal. Trio learning: a new strategy for building hybrid neural trees,Neural Networks for Computing (Snowbird), 1993, and published inInt. Journ. of Neural Systems, vol. 5, pp. 259–274, 1994.

    Google Scholar 

  9. F. d'Alché-Buc, J.-P. Nadal, D. Zwierski. Hybrid trees for supervised learning: towards extraction of decision rules,Proc. ECAI'94 Workshop on Combining Symbolic and Connectionist Processing (Amsterdam), pp. 133–142, 1994.

  10. M. I. Jordan, R. A. Jacobs. Hierarchical mixtures of experts and the EM algorithm,Neural Networks for Computing (Snowbird), 1993, andNeural Computation, vol. 6, pp. 181–214, 1994.

    Google Scholar 

  11. S. I. Amari, N. Murata. Statistical theory of learning curves under entropic loss criterion,Neural Computation, vol. 5, pp. 140–153, 1993.

    Google Scholar 

  12. J. Rissanen,Stochastic complexity in statistical inquiry, World Scientific, Teaneck, New Jersey, 1989.

    Google Scholar 

Download references

Author information

Authors and Affiliations

  1. Laboratoires d'Electronique Philips, 22 av. Descartes, B.P. 15, F-94453, Limeil-Brévannes Cedex, France

    Florence d'Alché-Buc & Jean-Pierre Nadal

  2. Laboratoire de Physique Statistique, Ecole Normale Supérieure 24 rue Lhomond, F-75231, Paris Cedex 05, France

    Jean-Pierre Nadal

Authors
  1. Florence d'Alché-Buc
    View author publications

    You can also search for this author in PubMed Google Scholar

  2. Jean-Pierre Nadal
    View author publications

    You can also search for this author in PubMed Google Scholar

Rights and permissions

Reprints and permissions

About this article

Cite this article

d'Alché-Buc, F., Nadal, JP. Asymptotic performances of a constructive algorithm. Neural Process Lett 2, 1–4 (1995). https://doi.org/10.1007/BF02312347

Download citation

  • Issue Date:

  • DOI: https://doi.org/10.1007/BF02312347

Keywords

Search

Navigation