Abstract
In this paper we propose a practical mechanism for extracting information directly from the weights of a reference artificial neural network (ANN). We use this information to train a structurally identical ANN that has some variations of the global transformation input-output function. To be able to fulfill our goal, we reduce the reference network weights by a scaling factor. The evaluation of the computing effort involved in the retraining of some ANNs shows us that a good choice for the scaling factor can substantially reduce the number of training cycles independent of the learning methods. The retraining mechanism is analyzed for the feedforward ANNs with two inputs and one output.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Hagan, M.T., Demuth, H.B., Beale, M.: Neural Networks Design, MA: PWS Publishing, Boston (1996)
Hassoun, M.H.: Fundamentals of Artificial Neural Network, MA: MIT Press, Cambridge (1995)
Nastac, D.I.: Contributions in Technical Systems Quality Modelling through the Artificial Intelligence Methods, Ph.D. dissertation, Polytechnic University of Bucharest (2000)
Zhang, Y., Peng, P.Y., Jiang, Z.P.: Stable neural controller design for unknown nonlinear systems using backstepping, IEEE Trans. Neural Networks 6 (2000) 1347–1360
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 2003 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Nastac, DI., Matei, R. (2003). Fast Retraining of Artificial Neural Networks. In: Wang, G., Liu, Q., Yao, Y., Skowron, A. (eds) Rough Sets, Fuzzy Sets, Data Mining, and Granular Computing. RSFDGrC 2003. Lecture Notes in Computer Science(), vol 2639. Springer, Berlin, Heidelberg. https://doi.org/10.1007/3-540-39205-X_77
Download citation
DOI: https://doi.org/10.1007/3-540-39205-X_77
Published:
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-14040-5
Online ISBN: 978-3-540-39205-7
eBook Packages: Springer Book Archive