The influence of a classifiers' diversity on the quality of weighted aging ensemble

Abstract

The paper deals with the problem of data stream classification. In the previous works we proposed the WAE (Weighted Aging Ensemble) algorithm which may change the line-up of the classifier committee dynamically according to coming of new individual classifiers. The ensemble pruning method uses the diversity measure called the Generalized Diversity only. In this work we propose the modification of the WAE algorithm which applies the mentioned above pruning criterion by the linear combination of diversity measure and accuracy of the classifier ensemble. The proposed method was evaluated on the basis of computer experiments which were carried out on two benchmark databases. The main objective of the experiments was to answer the question if the chosen modified criterion based on the diversity measure and accuracy is an appropriate choice to prune the classifier ensemble dedicated to data stream classification task. © 2014 Springer International Publishing Switzerland.