Neural network research on the digital transformation of industry

Subject. A study of the state of the industrial sector of the economy of the regions of the Russian Federation in order to enhance digital transformation as one of the innovative tools that determine sustainable development and technological leadership of the State. Objectives. Solving the multi–criteria problem of analyzing the development of the digital transformation of industry in the subjects of the Russian Federation, which is a task characterized by non-trivial formalization by means of the proposed modern practical method – cluster data analysis based on neural network modeling. 85 out of 89 regions of Russia were selected as research objects, qualified by 14 official criteria presented on the website of the Federal State Statistics Service. An overview of scientific publications by Russian and foreign authors is provided from the perspective of priority approaches to the development of digital transformation of the real sector of the economy. In order to promote digital transformation, dynamic indicators of the use of digital technologies and the use of industrial robots by organizations within the federal districts of Russia are presented. Methods. To study the multifaceted statistical parameters, the method of neural network cluster analysis using information technology is used. The clustering method through neural network modeling is not burdened with restrictive barriers. At the same time, there is no interference from the external environment in the work of the self-organizing artificial neural network. This method provides a visual representation on the plane of the results of neural network cluster analysis. Results. As a result of clusterization, the regions of the Russian Federation were divided into eight cluster formations. The hypothesis put forward in the work about the discrepancy between the structure of clusters and the structure of the federal districts of the Russian Federation is confirmed. A significantly different volume of clusters was obtained: the amplitude of the change in the number of subjects in the clusters is 20. A diverse level of development of the digital transformation of industry has been identified in accordance with the considered indicators on a cluster scale. Conclusions. The application of a progressive research method for the digital transformation of the industry of the Russian Federation, created on the basis of neural network modeling and information technology, is presented. The results of the work made it possible to assess the state of digital transformation in the real sector of the economy among the major challenges from external sources. When solving tasks to promote the strengthening of the country's technological leadership in the context of national goals, it is necessary to use various core areas of digital transformation of the industrial sector in the constituent entities of the Russian Federation, taking into account their specifics in the space of cluster formations.