Combination Model Of ARIMA And GM In Forecasting The Levels Of Urban Residents Consumption

China as a big country with growing economical,its urbanization is deepening.The different levels of consumption of residents are various in areas.The indicators react the people's living.Consumption is a part of domestic demands,at the same time it can improve the economic growth.So this article shows the meaning of the predication of consumption of urban residents of shaanxi province.The paper use different mathematics models to combine and predicate the levels of urban residents consumption.The consumption data resource are from the shaanxi bureau of statics.the paper start to obvious the character of the data distribution,use the time sequence diagram to analysis,then use the obvious result of The ARIMA(p,d,q)model and GM(1,1)model to group and predict.Take use of arithmetic average method to find out the coefficient.take The error sum of squares(SSE)and Mean absolute error(MAE)to inspect the results of the combination and simulation.At the same time,compare the result of two models and their result.finally chose the best model of predication.Although It is unable to predict urban residents consumption level and the exact number of changes in Shaanxi Province,the prediction results were still is based on the historical trend of the time sequence,so the results is more accurate,which can provide a reference for individuals or departments that needs the conclusion of this article.