A Small Sample Of Time Series Analysis Of Per Capita GDP

Time series analysis methods of modern data processing is widely used in practical engineering. It requires the number of the datas which are observed is not less than 50. However, in practical applications, the number of the datas which are observed is not allow to be large for various reasons,, and sometimes only a dozen in this case, we need to study the small sample of time series modeling methods in this case.In recent years, with the development of our economic, per capita GDP has a significantly growth, but there are many factors which impact the per capita gross domestic product and the changes are difficult to grasp, that brings difficulties to the forecasts. This paper attempts to do a trial discussion for the prediction of per capita gross national product in the harsh condition of a small sample with the method of time-series analysis. (The datas of this article are come from the per capita gross national product of the years 1988-2007, the datas of the years 1988-2005 are prepared for modeling, and the datas of the year 2006 and 2007 are prepared for the comparison with the data of forecast).At the same time, the datas of GDP from 1996 to 2005 are analysized in the paper. Based on Grey Theory, a Grey Model was constructed to forecast GDP and its precision was highly relative. And with the use of the grey incidence degree, the influent degree of the first, second and third industry of China on GDP can be analyzed efficiently.