Functional Networks and Analysis of Variance for Feature Selection
In this paper a method for feature selection based on analysis of variance and using functional networks as induction algorithm is presented. It follows a backward selection search, but several features are discarded in the same step. The method proposed is compared with two SVM based methods, obtaining a smaller set of features with a similar accuracy.
KeywordsSupport Vector Machine Feature Selection Mean Square Error Feature Selection Method Functional Network
Unable to display preview. Download preview PDF.
- Castillo, E., Cobo, A., Gutiérrez, J.M., Pruneda, E.: Functional Networks with Applications. Kluwer Academic Publishers, Dordrecht (1998)Google Scholar
- Sanchez-Maroño, N., Alonso-Betanzos, A., Castillo, E.: A new wrapper method for feature subset selection. In: Proc. European Symp. on Artificial Neural Networks, pp. 515–520 (2005)Google Scholar
- Bishop, C.: Neural Networks for Patter Recognition. Oxford University Press, New York (1995)Google Scholar
- Blake, C., Merz, C.: UCI repository of machine learning databases (1998), http://www.ics.uci.edu/mlearn/MLRepository.html