Application Of Multivariate Statistical Analysis In Evaluation Of Regional Economy

Evaluational result of regional economic development level is an important prove for governmental sectors to judge the present development level and plan the strategies of this area. There are multivariate statistical analysis,linear programming and decision theory in the mathematic mothod we often used to analysis regional economy at present. Multivariate statistical analysis includes principal component analysis,factor analysis,cluster analysis and regression analysis, while usually only one of these methods is used in the evaluation, which exist some limitations that the differentes of analysis leads to the inconsistant of the result of the same group of data for different methods.Wuhan Metroplitan Area which plays an inportant role in comprehensive reform of the two-oriented social construction and coordinated development of regional economy in our country, is the leading role of National strategy of the rising of central China. And it takes a radial effect on other areas in Hubei Province.Taking Wuhan metroplitan area as the research object, this paper selects 30 economic indicators through constructing the evaluation index system which reflects the comprehensive economic strength of each city. Analyze the data by using SAS software as well as the principal component analysis method, the factor analysis method,and check the consistency of the results through concordant coemcients kendall test, then combine those two results.Finally,classify the 9 cities on the basis of the results with cluster analysis, and put some suggestions about the economic development of Wuhan metroplitan area.The article concluded a better analytic method of the regional economic development through the analysis of specific case.This method can aviod the inconsistency of different Multivariate statistical analysis and approaches to the same statistic analytic result.In the other words, we should first use many methods to analyse the same quota statistics, and then examine the consistency of those results, under the good consisitency, we can combine the results and get the final analytic result.