Research On Analysis And Forecast Of The Automobile Production In China

The automobile industry is an important part of the national economy.At the same time,the development of the automobile industry drives machine,electron,iron and steel industry,etc.In recent years,the automobile industry has gotten swift and violent progress.Accurate prediction of development trend of the automobile production in China provides data support for the balance between production and marketing in automobile market,it offers effective guidance to our government and relevant departments that help to make important decisions.The significance of time series analysis lies in studying a certain time series in long-term series in the form of statistical regularity.It is usually considered that there are 3 purposes of time series analysis:considering the dynamic systems,predicting the future events and controlling events in the future through the intervention.In this thesis,the data come from the National Bureau of Statistic of the People’s Republic of China web site,this study uses the data from January 2001 to December 2015.Because of the lack of some data,we have to fill missing values of the automobile production using R language before establishing a prediction model.Holt-Winters exponential smoothing method,ARIMA model and BP network are used based on time series.Exponential smoothing method is a kind of empirical analysis,which needs no requirement for data distribution as well as easily operates.ARIMA model focuses on the probability of time series itself or random nature,using the AIC criterion to select the optimal estimation model.BP network possesses nonlinear processing ability and self-learning ability.The predicted values are consistent with the trend of the true values of the three methods and the relative errors are within a reasonable range.The BP neural network prediction results are the most ideal.