Improving Of BP Neural Network And It's Applications

Neural network is attended by people because of it'characteristics The result of research has indicated that neural network is more effect than traditional method in dealing with information of instinct and thought. It will develop well. BP neural network includes the essential part of neural network theory, it is applied widely because of it's simple structure. it can solve the complicated problem of non-linear function.This paper introduces the basic theory of neural network, including the defining of neural network, the method of studying and appraising many neural network model first.then introduces BP neural network model, in particular, introduces the model.at the same time, it appraises it's algorithmic and introduces some improving. Last, this paper brings forward improving and puts it into practice of the positive research to family's disposable income with the urban residents of during the State's Tenth Five-Year Plan period of "the Nineth Five-Year Plan". we can get that the improving method is better than the traditioal method.The throught compensating through the error finally is proposed by a kind of association model based on dispersed grey and neual network,and carried on the positive research to family's disposable income with the urban residents of during the State's Tenth Five-Year Plan period of "the Nineth Five-Year Plan".The result shows that combination model has the better precision of prediction than the single model, so there will be deeper development space to combination model.