Author(s)

I. Gertsbakh & I. Yatskiv

Abstract

The purpose of dynamic classification in application to the economic development of EU countries is to work out an Economic Welfare Growth Index (EWGI), the use of which would make it possible to establish, for each year and each country in the study, the level of its economic welfare, to estimate the rate of advance of each country toward economic well-being, to compare different countries and to make forecasts. EWGI is defined as an \“optimal” linear combination of leading socio-economic parameters, such as GDP, CPI (corruption transparency index), level of unemployment, inflation, average life expectancy. On the first stage we consider one \“training” year and classify all EU countries by using standard clustering algorithms to single out two groups: P and R, having the \“worst” and the \“best” values of the parameters, respectively (\“poor” and \“rich”). On the second stage we use these groups for constructing the Fisher Discriminant Function (FDF). For sake of convenience the FDF is linearly transformed in such a way that the centers of P and R give the FDF values 0 and 10, respectively. This transformed FDF is taken as the EWGI. It has range [-2, 4] for P, [9, 14] for R and [4, 9] for countries in the \“middle”. The final output of our analysis is a collection of time series (graphs) representing the dynamics and the behaviour pattern of EWGI for each country in the period 1995-2004. Similar approach can be used also in other areas, such as transportation, education, health care, whenever the development in time is described by a multidimensional time series. Keywords: economic welfare index, cluster analysis, Fisher discriminant function.

Keywords

economic welfare index, cluster analysis, Fisher discriminant function.