Construction And Application Of Grey Forecast System Model

This paper describes the forecast for the importance and significance of financial investments,description of the value of the Grey Prediction. Followed by the development of the theory of greysystem are stated in the origin and the research direction, its basic meanings and content, the greysystem model generated sequence and the five step modeling ideas and its construction and five-step modeling idea of Grey System Model. Gray system model which describes the three mainGM (1,1) model construction and application, and examples of modeling and analysis simulationaccuracy of the three models. Then select the simulation model for high precision index ShanghaiComposite Index closing price modeling predicts, we found that GM (11) model of the Shanghaiindex prediction error is bigger, then use residual GM (1,1) model to modify the simulationprecision of exponential model, although it adjusted the accuracy, but the error is bigger, so theprediction GM (1,1) is not applicable to the Shanghai index’s closing price.Then, using GM (1,1) model of three models: old information model, new information modeland metabolism model, taking WISCO shares (60005) respectively closing price data forsimulation and prediction, and compare the prediction accuracy of the three models and found thatthe metabolic model prediction error is minimized. Therefore, we pick WISCO shares metabolicmodel from January2010closing price of4to March25,2014date to predict the trading day,however, on the basis of the original prediction model of metabolism were expanded and applied,and model of the predicted accuracy is higher.Finally, according to the grey system theory and grey forecasting the characteristics of thesample data, this article selects daily closing price of stock index futures to forecast, we put theoriginal GM (1,1) model and the improved GM (1,1) model to predict the price of future stockindex and compared stock index futures closing price with the traditional grey prediction model’s.We use Matlab to achieve modeling process, the precision of prediction and test, and compared thecontrast images of the predicted closing price and the actual closing price of two models, andanalyze the accuracy of model predictions. In this paper, with the previous forecast of gray model is somewhat different: first, toimprove the prediction method based on the original gray prediction method based on grayprediction obtained algebraic recursive equation models. Second, we used the program to predictthe dynamic metabolic model, a large number of historical transaction data was extracting directlyby the program to get the actual comparison chart data and forecast data,this selection of data isalso constantly updated, guarantee the timeliness of data, thus improves the prediction precision,the GM (1,1) model forecast more meaningful.