Design Of Multiple Classifier Systems Based On Evidential Neural Network

To make full use of the data information and improve the classification performance, a new evidential neural network classifier is proposed and a novel implementation of multiple classifier systems based on the new evidential neural network classifier is presented in this paper. The ambiguous data contained in the training data is considered as a new class - compound class and the training data is reconstructed into a new data set with compound classes. The evidential neural network member classifiers are trained using the new data set with compound classes, and the classification output is modeled with the belief function. We use of a variety of fusion rules in evidence theory to combine the member classifiers. According to the experimental results on the artificial data sets and some UCI data sets, the new multiple classifier systems can effectively improve the classification accuracy compared with some other existing ways to implement multiple classifier systems.