Study of ensemble method of classifiers for neural networks based on K-means clustering

Aiming at diversity being a necessary condition of the ensemble learning, we study method for improving diversity of the neural networks ensemble based on K-means clustering technique. In this paper, we propose a selecting approach that is first to train many classifiers through training set with neural network algorithm, and to classify data on validation set using classifiers. And then we use the K-means algorithm to clustering the results of classifiers and select a classifier model from every cluster to make up of the membership of the ensemble learning. Finally, we study the performance of ensemble method by using vote fused method and compare performance with bagging and adaboost methods.