Comparative analysis of the Russian regions using machine learning methods for a set of indicators of electronic services

Abstract

Digitalisation is important not only for the work of professionals in their workplaces, but also for the well-being of specialists and adaptation to changes in the information sphere as well as for providing high-quality educational, medical and social services to the population. The paper uses machine learning methods to classify regions according to a set of indicators of electronic services and services. A clear division of the regions into two large, almost equal groups according to this set of indicators has been obtained. The use of well-known statistical criteria has demonstrated the statistical significance of such a division. Scattering diagrams are constructed as an example of the relationship of such indicators. The multiple correlation coefficient between the indicators of electronic services and services is 0.71, which indicates a close relationship between the indicators of digitalisation of services. In addition, the division of regions into clusters has been obtained using hierarchical clustering, which implies that Moscow has significantly overtaken other regions of Russia in providing electronic services and services, and the remaining regions are heterogeneous in this indicator, considering the considered indicators of digitalisation of the subjects according to the Federal State Statistics Service data for May 2024.

Keywords

• machine learning methods
• digitalisation
• statistics
• regression
• correlation
• clustering
• dendrogram
• classification of regions
• electronic services