Effect of Feature Selection on Bagging Classifiers Based on Kernel Density Estimators
Abstract
A combination of classification rules (classifiers) is known as an Ensemble, and in general it is more accurate than the individual classifiers used to build it. One method to construct an Ensemble is Bagging introduced by Breiman (1996). This method relies on resampling techniques to obtain different training sets for each of the classifiers. Previous work has shown that Bagging is very effective for unstable classifiers. In this paper we present some results in application of Bagging to classifiers where the class conditional density is estimated using kernel density estimators. The effect of feature selection in bagging is also considered.
Type
Publication
In K. Jajuga, A. Sokolowski, and H.-H. Bock (eds). Classification, Clustering and Data Analysis. Springer, Heidelberg, Germany
