Modeling And Prediction For Home-textile Price Index Based On Time Series

This paper takes all actual data of all sorts of home-textile price index as the researchbackground, which is from the website named home-textile industry of China. The relevanttheoretical methods and modeling forecasting methods of time series are applied to thehome-textile fabric price index fluctuation analysis.By analyzing the basic theory of constructing price index, Fischer ideal index is chosento achieve the price index of home-textile fabrics. The statements are made on the formedprice index at all levels after classification. The features of the auto regressive model, themoving average model, the auto regressive moving average model and the auto regressiveconditional heteroskedastic model which are classic models in time series are combined.And the models of the home-textile price index are built through the stationarytransformation, heteroscedasticity effects testing, and model-order determination and so on.The results show that the fitting curve has the same trend with the real curve, both thefitting and prediction results are in the theoretical confidence interval, forecast accuracy ishigh, error fluctuation is small, assessment and prediction can be achieved on the relevantproduct market in this method. Different models are built which are based on the methodabove. In the case of different sample sizes, the relationship between forecast accuracy andsample size is analyzed.10points,5points and3points are predicted based on the samesample size. The result shows that the forecast accuracy is the highest when3points arepredicted. So the method of building models of the time series is more suitable forshort-term forecasts. The models of the four sets and six sets of the home textile are builtbased the same modeling method. The contact and the difference between the finished product and the home-textile fabric price index are analyzed. Taking price index in ChinaTextile Economic Information Network of raw materials as the study background, thispaper uses the same modeling approach to build the models of the raw material price indexsuch as cotton, rayon and polyester.This paper predict the operation situation of the home-textile fabric price index.Thepredictions are ideal which are matched to the actual situation.Time-series method can beapplied to short-term macroeconomic forecasting home-textile market and it can providereferences of the price alteration to the producers and market operators and managers in thehome-textile industry.