Abstract
Neural network has a wide range of practical background, wherever information processing, automation, engineering or economy, medical, we will encounter. It is a information processing systems which is designed to mimic the structure and function of human brain. It has some features and functions of human brain. BP, which is one of the important algorithm of neural network, plays a special roal in prediction. BP is a multi-layer perceptron, which is based on back-propagation. The basic idea is that: first of all, getting the errors through forward-propagation of input signal, Then, propagating the errors back to units of every layer. At last revising the weight of units according to the error signals. The above process will be repeated until the errors reduced to an acceptable level. The paper is mainly about the basic principles of neural network and BP network, and achieves three specific applications of BP network, solve these problems through programming by using MATLAB.
Keywords: neural network; BP network; the Matlab; Specific application
