A Method for Building Partially Connected Neural Network

Abstract

This paper focuses mainly on application of Partial Connected Back Propagation Neural Network (PCBP) instead of typical fully connected neural network (FCBP), as  PCBP with less connections learns faster than FCBP. The initial neural network is fully connected, after training with sample data, a clustering method is employed to cluster weights between input to hidden layer and from hidden to output layer, and connections that are relatively unnecessary are deleted, thus the initial network becomes a PCBP network. PCBP can be used in prediction or data mining by training it with data that comes from database. At the end of this paper, several experiments are conducted to illustrate the effects of PCBP using the submersible pump repair data set.