Improving the combination module with a neural network

Abstract

In this paper we propose two versions of Stacked Generalization as the combination module of an ensemble of neural networks. The first version only uses the information provided by expert networks. The second one uses the information provided by experts and the input data of the pattern that is being classified. Finally, we have performed a comparison among 6 classical combination methods and the two versions of Stacked Generalization in order to get the best method. The results show that the methods based on Stacked Generalization are better than classical combination methods. © Springer-Verlag Berlin Heidelberg 2006.