An information fusion approach and its application based on D-S evidence theory and neural network

This paper puts forward a multisensor information fusion method, combining the BP neural network with the D-S evidence theory, which not only gets over the shortcoming of not being real-time caused by unitary neural network that reaches high accuracy at the cost of many times of iteration, but trains the neural network by using large amount of standard samples, so as to make the D-S evidence theorypsilas decision on the system become more accurate. Experiment shows that the method mentioned above has broad prospect of being applied in the field of intelligent medical diagnosis.