| Stock is the product of market economy, stock trading and distribution will promote the development of market economy, has experienced 10 years of development, China’s stock market has never mature to mature gradually, and the stock market is an important part of financial market, so the stock market in our country has become the most important capital markets. Variation trend of stock prices is a national political, economic, and comprehensive reflection of life situations. In recent years, due to the influence of international environment, China’s stock market’s ups and downs, it also brought a lot of confusion to investors, how to accurately analyze the stock market and the stock price fluctuation trend has become an important topic, also is a difficult task, for the nation to economic macro-control and taken investors how to make optimal decisions, it is of great significance.In this paper the related theory of time series has carried on the simple introduction, for each prediction model has carried on the detailed research, In this paper, we obtain the following results:(1) Time series and the stationary time series were introduced simply, the ARMA model is introduced in traditional AR and MA models and introduces in detail the basic principle of ARMA model and the modeling process, and analyses the clad steel shares in the stock and forecast, prove the validity of the results.(2) The definition of neural network theory has carried on the simple introduction, then mainly introduces the structure of the BP neural network and BP algorithm design, and finally using BP algorithm to baotou steel co shares in the stock’s closing price of the fitting and forecasting, and the error is analyzed.(3) By using ARMA model and BP neural network model combined into BP-ARMA model, is applied to the prediction of stock, and compares the results of prediction and analysis, the experimental results show that the combined model prediction accuracy was significantly higher. |
PDF Full Text Request