Statistical Methods Of Research, Divided Into Provinces Of China's Gdp-based Economic Region

Data - grouping is an important problem in statistics because of its wide applications and complexity. According to given GDP data, the paper divides our country into several economic zones. They exist objectively. The division of economic areas is helpful to integrally take advantage of the nature, society and economy of each area, to organize reasonably the economic relationship, capital investment and policy establishment within all the economic areas. It is also helpful to promote the continuous and harmonious development of our national economy by dividing the work within all the areas.Many statisticians and economists have played more attention on the importance of dividing economic areas. They gradually introduce statistical methods into the division of areas, for example, clustering analysis and principal components analysis etc. The two methods and the method of measuring difference between areas are given in document [12]. Document [14] proposes the modified clustering analysis: variance square sum clustering analysis, weighted variance square sum clustering analysis etc. But they have not explained the number of the division of economic zones, and only show how to divide the economic zones.This paper solves the problem of dividing the districts into different economic zones according to the GDP data of our country. The three main results of our paper are followed as1. Introducing AIC(Akaike Information Criterion).Akaike proposed AIC in 1973. It is widely used in many fields of statistics and has good results in practice. This paper introduces AIC which is the basis of this paper.2. Applying AIC to data-grouping and giving a formula that shows how to group the data. Furthermore, we give a new method of grouping according to the thought of AIC, and name it as variance criterion.3. Based on the GDP data, we give the results of grouping, so we solve the practical problem.