The Research Of BP Neural Network Model In Stock Price Forecasting

With the economic growth and the conversion of people's investment consciousness, stock has become an important part of people's life in modern time. The stock investment has become a important way of the numerous families and the individual managing finances, which decides a family property income. The proceeds of stock investment always equal the risk. So establishing a stock forecasting model, which has higher operation rate and precision, has theoretical significance and applicable value.This dissertation compares various stock forecasting methods. Provide operational support to the stock this topic conducted in-depth exploration and research is proposed in the paper.Based on studying these existing problems of BP algorithms in stock forecasting, including the slow learning speed, local minimum and the low prediction, an improved BP neural network algorithm is presented. This algorithm enhances the convergence speed by using additional momentum item and taking auto-adapted study rate measure. The selection of the BP neural network training sample has the tremendous influence to the network pan-ability, how to select the appropriate training sample from the complex sampled data is a difficult question. This method further enhance the network pan-ability and the forecast precision.According to the principle of stock prediction based on BP network, the prediction model of stock has been established. The stock is predicted by adopting improved BP algorithm and simulation experiments are conducted through MATLAB.At last taking the stock price of 000698 for example, the established prediction model is trained and then its stock dada are predicted using the trained network and good effect has been grained.