Research on Structure and Application of Superior Combination Forecasting Model
As the stock prices are random walk, so the pricing in the securities of RBF neural network model, gray GM (1,1) model, ARIMA model don't have the timeliness. Through a comprehensive analysis of these three models, this paper proposed an optimal combination forecasting model, which combined with the a-bove three models in the collection of useful information. On this basis, the Shenzhen Development Bank a stock's closing price in the year 2007 was selected as the research sample, on which an empirical analysis was made by using the four models. The results show that this method to predict the securities prices has a good prediction accuracy and high reliability.
RBF neural network modelGM(1,1) modelARIMA modelcombination forecasting model
国家科技期刊平台
NSTL
万方数据
维普
