The Application Of BP Neural Network In Predicting Stock Market

Stock market is an arena filled with chance and trap. Since1990, when the stocks wasfor public offerings in Shanghai and Shenzhen, it becomes a common part of citizens’economic behavior to invest in stocks. Although stock would return investors with high profit,it also companies with high risk. Under such a circumstance, more and more investors andinstitutions begin to focus on the trend of stock market, trying to predict the trend of stockthrough large amount of stock data. It has both magnificent theory meaning and practicalvalue to research and predict inner rules of stock market successfully.This paper attempts to use neural network analysis and forecasting stock. The stockmarket is a very complex nonlinear dynamic system, the neural network has a strongnonlinear approximation ability and self-adaptive characteristics. By the finishing stock pricehistory data of the BP neural network training learning historical data, can effectively find thelaw of the changes in the market price of a stock, to predict stock price trends.This paper analyzes the principle of predicting stock market based on the BP neuralnetwork, and establishes three-layer feed forward neural network model. In the experiments,by adapting the parameters in the BP neural network, we could achieve a well result ofnetwork learning. Then, we got well result by testing five stocks from NASDAQ as examples,on which our forecast model was applied to predict the future trend.