Transferring neural network based knowledge into an exemplar-based learner (original) (raw)

Abstract

This paper investigates knowledge transfer from a neural network based system into an exemplar-based learning system. In order to examine the possibilities of such transfer, it proposes and evaluates a system that implements a collaborative scheme, where a particular type of neural network induced by the neural system RuleNet is used by an exemplar-based system (NGE) to carry on a learning task. The proposed collaboration between the two learning models implemented as the hybrid system RuleNet→NGE is feasible due to the similarity of the concept description languages employed by both. The paper also describes a few experiments conducted; results show that the RuleNet-NGE collaboration is plausible and, in some domains, it improves the performance of NGE on its own.

Access this article

Log in via an institution

Subscribe and save

Buy Now

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

Instant access to the full article PDF.

Similar content being viewed by others

References

  1. Mitchell TM (1997) Machine learning. McGraw-Hill, New York
    MATH Google Scholar
  2. Tschichold-Gürman NN (1995) RuleNet: a knowledge-based neural network model with application examples in mobile robotics. Ph.D. Thesis, ETH, Zürich
  3. Tschichold-Gürman NN (1997) The neural network model RuleNet and its application to mobile robot navigation. Fuzzy Sets Syst 85:287–303
    Article Google Scholar
  4. Salzberg SL (1989) Learning with nested generalized exemplars. Ph. D. Thesis, Harvard University, Cambridge
  5. Salzberg SL (1991) A nearest hyperrectangle learning method. Mach Learn 6:252–276
    Google Scholar
  6. Nauck D, Klawonn F, Kruse R (1997) Foundations of neuro-fuzzy systems. Wiley, West Sussex
  7. Medin DL, Schaffer MM (1978) Context theory of classification learning. Psychol Rev 85(3):207–238
    Article Google Scholar
  8. Aha DW, Kibler D, Albert MK (1991) Instance-based learning algorithms. Mach Learn 6:37–66
    Google Scholar
  9. Cover T, Hart P (1967) Nearest neighbor pattern classification. IEEE Trans Inf Theory 13:21–27
    Article MATH Google Scholar
  10. Aha DW (ed) (1997) Lazy learning. Springer, Heidelberg
  11. Wettschereck D, Dietterich TG (1995) An experimental comparison of the nearest-neighbor and nearest-hyperrectangle algorithms. Mach Learn 19:5–27
    Google Scholar
  12. Hettich S, Blake CL, Merz CJ (1998) UCI Repository of machine learning databases. Department of Information and Computer Science, University of California, Irvine (http://www.ics.uci.edu/∼mlearn/MLRepository.html)
  13. Nicoletti MC, Figueira LB, Ramer A (2004) Transferring symbolic knowledge into a RuleNet neural network. Proceedings of international conference on computer cybernetics, Vienna, pp 175–182
  14. Nicoletti MC, Figueira LB, Ramer A (2005) Utilizing previously induced knowledge to initialize the learning process of a neural network. In: Elmereich W, Machado JT Rudas IJ (eds) Intelligent systems at the service of mankind, vol. II. UBooks, pp 365–376
  15. Nicoletti MC, Figueira LB, Hruschka ER Jr (2005) Initializing an exemplar based learning process from a RuleNet network, In: Proceedings of the hybrid intelligent systems (HIS’05), pp 125–130

Download references

Acknowledgments

To CNPq, CAPES and FAPESP for the support provided to the first, second and third author of this paper, respectively. Thanks to Leonie C. Pearson for numerous insightful comments and suggestions and for proof reading this paper. This paper is an extended version of an earlier conference paper [15].

Author information

Authors and Affiliations

  1. Computer Science Department, UFSCar, Sao Carlos, Brazil
    Maria do Carmo Nicoletti & Estevam R. Hruschka Jr
  2. Physics and Math Department, DFM-FFCLRP, University of Sao Paulo, Sao Carlos, Brazil
    Lucas Baggio Figueira

Authors

  1. Maria do Carmo Nicoletti
  2. Lucas Baggio Figueira
  3. Estevam R. Hruschka Jr

Corresponding author

Correspondence toMaria do Carmo Nicoletti.

Rights and permissions

About this article

Cite this article

do Carmo Nicoletti, M., Figueira, L.B. & Hruschka, E.R. Transferring neural network based knowledge into an exemplar-based learner.Neural Comput & Applic 16, 257–265 (2007). https://doi.org/10.1007/s00521-007-0088-8

Download citation

Keywords