Methods of cluster analysis in regional studies

Abstract

The paper proposes a technology for cluster analysis of the constituent entities of the Russian Federation which can become the basis for a methodology for developing targeted support measures from the state for the purpose of regional development. A cluster analysis of the constituent entities of the Russian Federation was carried out for a number of economic indicators characterizing the sectoral orientation of the regions. Two approaches were used: k-means clustering (with visualization using the t-SNE method) and hierarchical clustering. A comparative analysis of the obtained results of clustering the economic space of the Russian Federation was carried out.