Investment attractiveness estimation of Russian mountain territories using artificial intelligence

Abstract

Introduction. At the current stage of Russia’s development, the need for investment inflows into the mountainous areas of our country has increased. Therefore, effective tools are required to monitor not only the current situation, but also to forecast its changes in the future. The purpose of the article is to assess the investment attractiveness of the mountainous areas of the Russian Federation using artificial neural networks and their clustering by the level of investment attractiveness, which made it possible to develop short-term investment forecasts. Materials and methods. The information base of the study was the statistical data of Rosstat and EMISS for 2019–2024. The work uses a modified author’s approach to assessing the investment attractiveness of the mountainous areas of Russia based on the index method. Clustering of regions is carried out using the method of self-organizing maps of T. Kohonen. The main features of the applied research methodology are the simultaneous use of the index method in combination with artificial intelligence (artificial neural networks). Neuro-modeling is carried out in the demo version (with a limitation on the data array – no more than 150 observations) of Deductor Studio Lite 5.1. Results and discussion. As a result of ranking the investment attractiveness of mountainous territories of Russia for 2019–2023, a number of conclusions were made: 1) the stable leader of the rating since 2020 is the Krasnodar Territory. In addition, the group of leaders of the rating on a regular basis included the Chelyabinsk, Sverdlovsk and Tyumen regions, other regions; 2) the stable outsider of the rating is the Republic of Tyva. In addition, the Republic of Ingushetia, the Karachay-Cherkess Republic, the Chechen Republic, other regions; 3) all regions were grouped into several clusters according to the stability of the results of the retrospective assessment and the direction of change in the place in the ranking, where regions with a relatively stable place in the ranking (conditionally stable position) and subjects of the Russian Federation with progressive (jumpy) multidirectional movement (up and down the ranking) were identified. On this basis, a short-term forecast of the investment attractiveness of mountainous territories was compiled using artificial neural networks. Conclusions. The article proposes and tests the author’s approach based on a combination of several methods of assessment, clustering and forecasting the investment attractiveness of mountainous territories of Russia, which made it possible to compile a rating of investment attractiveness of regions, identify stability zones of the results of retrospective assessment, visualize calculations using cartographic and graphical methods and give a short-term forecast of investment development of mountainous territories using artificial neural networks. Proposals for practical application and directions for future research. The rating of investment attractiveness of mountainous territories of Russia showed regions that need increased attention from the federal center for the successful implementation of the strategy of spatial development of Russia. Directions for future research – search for investment opportunities for sustainable tourism development.