Time Series Analysis Of The Balance Of Savings Deposits Of Residents In Shandong Province

This paper we use multiplicative seasonal model, season index analysis model and the auto-regressive model respectively to fit and analyze the balance of saving deposits of residents of Shandong Province. And we will compare the test results of the three models to select the optimal one to predict the balance of saving deposits of residents of Shandong Province monthly and carry on the analysis according to the prediction results.Firstly we analyze the monthly data of January2001to December2012. According to the timing sequence diagram we found that the data has a in-creasing trend with time and seasonal regularity to a certain extent. After one-order difference the sequence tends to be stable, but the autocorrelation and partial autocorrelation coefficient graphics show seasonal effect to a certain extend. Hence we carry on the difference of the12step to one-order difference sequence. Finally we use the multiplicative seasonal model to fit and forecast the data. Then we use the model to predict the value of each month of the year of2013and calculate their relative errors by the observe values. Results show that the absolute value of relative errors are all within2%.As a comparison, we use the seasonal index analysis model and auto-regressive model respectively to fit and analyze the original sequences. By the models which go through the test, we predict the the value of each month of the year of2013and calculate their relative errors of the two models by the observe values. According to the results, the prediction of the multiplicative seasonal model are most accurate; next is the seasonal index analysis model whose absolute value of relative errors are within3%; the auto-regressive model is worst whose absolute value of relative errors are within7%.So we choose the multiplicative seasonal model to predict the balance of saving deposits of residents of Shandong province in each month of the year of2014.