Constructing Social And Economic Indicators For EU Countries Using Dynamic Classification: Case Studies
Free (open access)
153 - 161
I. Gertsbakh, I. Yatskiv & O. Platonova
This paper presents applications of the dynamic classification algorithm (DCA) described in Gertsbakh and Yatskiv (Dynamic Classification: Economic Welfare Growth in EU During 1995-2004. Proceedings of International Conference \“Data Mining 2006”, 11-13 July 2006, Prague. WIT Press. 2006, p.53-62) to the development of three socio-economic indicators: National Health Index (NHI), Population Mobility Index (PMI) and Logistic Performance Index (LPI). In each of these three cases we work with the respective data for EU-25 countries. The essence of the DCA is to transform a multidimensional vector to a scalar on the basis of combining cluster and discriminant analysis. In Gertsbakh and Yatskiv, the DCA was used for obtaining a measure of socio-economic welfare and its dynamics for EU-25 countries over the period 1995-2004. The output of DCA is a collection of time series (graphs) representing the dynamics and the pattern of the scalar socio-economic indicator for each country in the considered time period. Keywords: cluster analysis, Fisher discriminant function, time series, socioeconomic indicators. 1 Introduction: the data and the algorithm The data. The row data are given as a three-dimensional array of entries having the following form: )], ( ),..., ( ),..., ( [ ) ; ,..., 1 ; ( 1 t x t x t x t k j i ik ij i = = y (1)
cluster analysis, Fisher discriminant function, time series, socioeconomic indicators.