Cluster analysis of international indicators of human development in the context of the religious factors based on Kohonen neural network

Abstract

Introduction. The essence of the study is to figure out how religious mental factors influence the socio-economic life, inhibit or accelerate economic development. Accordingly, the object of the study is to determine the characteristic features of the world's religions and their impact on the socio-economic development of countries.

Purpose. The aim of the research is to verify on the basis of current data of human development indicators a working hypothesis that the vast majority of developed countries confess Protestantism, as well as all the Protestant countries are highly developed, by applying data mining techniques, in particular cluster analysis as an effective tool for classification.

Methods. The method of correlation analysis has been applied to test the input data, and the clustering objects method using Kohonen self-organizing maps based on single-layer neural network of direct distribution has been used.

Results. As a result of the cluster analysis we have grouped a lot of countries (clusters) using similar indicators of the development. A number of clusters according to the Kohonen algorithm, which is applied in the research, is determined automatically, therefore, there were highlighted six clusters. The cluster of most developed countries includes 17 countries, mostly Protestant, except Japan with its own national religion, and two Orthodox countries: Greece and Cyprus.

Analysis of the religious factors influence on human development indicators suggests that Protestant, Judaic and Confucian societies have achieve greater economic development than Catholic, Muslim and Orthodox, because the first ones are generally characterized by the same values. Although, undoubtedly, religion and other spiritual and mental characteristics are not the sole factors that determine economic behavior, but this study clearly shows that they have a dominant influence on the it.