Abstract
This paper presents a new method that can be used for symbolic knowledge extraction from neural networks, once they have been trained with the desired performance. Weights are the basis for this method. This method allows knowledge extraction from neural networks with continuous inputs and outputs, more precisely in problems dealing with the general linear regression model where exists multicolineality among the input and output. An example of the application is showed by comparison of the results between the regression and the neural networks results, concernig the estimation that gasoline yields from crudes. This example is based on detecting the most important variables when there exists multicolineality.
Supported by INTAS 952
This is a preview of subscription content, log in via an institution to check access.
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
Canavos, G.C.: Applied Probability and Statistical Methods. McGraw-Hill, Mexico (1988)
Castellanos, J., Castellanos, A., Manrique, D., Martínez, A.: A New Approach for Extracting Rules from a Trained Neural Network. In: Costa, E. (ed.) EPIA 1997. LNCS, vol. 1323, pp. 297–302. Springer, Heidelberg (1997)
Castellanos, J., Manrique, D., Martínez, A., Rios, J.: 4th World Congress on Expert Systems, Mexico, pp. 598–603 (1998)
Cuadras, C.M.: Métodos de Análisis Multivariante. Eunibar. (1981)
Chang, B.L., Hirsch, M.: Knowledge Acquisition and Knowledge Representation in a Rule-Based Expert System: Computers in Nursing, vol. 9(5), pp. 174–178 (1991)
Gail, A.C., Tan, A.: Rule Extraction: From Neural Architecture to Symbolic Representation. Connection Science 7(1) (1995)
Gallant, S.I.: Neural Networks, Learning and Expert Systems. MIT Press, Massachusetts (1993)
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1999 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Martínez, A., Castellanos, J., Hernández, C., de Mingo, F. (1999). Study of Weight Importance in Neural Networks Working with Colineal Variables in Regression Problems. In: Imam, I., Kodratoff, Y., El-Dessouki, A., Ali, M. (eds) Multiple Approaches to Intelligent Systems. IEA/AIE 1999. Lecture Notes in Computer Science(), vol 1611. Springer, Berlin, Heidelberg. https://doi.org/10.1007/978-3-540-48765-4_13
Download citation
DOI: https://doi.org/10.1007/978-3-540-48765-4_13
Publisher Name: Springer, Berlin, Heidelberg
Print ISBN: 978-3-540-66076-7
Online ISBN: 978-3-540-48765-4
eBook Packages: Springer Book Archive