Segment procedure neural networks

This paper deals with segment procedure neural networks and its learning algorithm. A novel procedure neural network is proposed based on procedure neurons. It aims at simulating the procedure of segment object programming and estimate system. Function approximation ability is discussed for segment procedure neural networks. It is indicated that this model is a generalized form of the known procedure neural networks, and it has all properties of the known procedure neural networks. This paper also presents learning algorithms for the segment procedure neural networks. An example dealing with high education grade system is proposed by the model and its efficiency is illustrated.