APPLICATION OF NEURAL NETWORKS IN SIMULATION OF CLUSTER-NETWORK RELATIONS IN OIL AND GAS INDUSTRY

Abstract

Despite the fact that the cluster approach is quite common in scientific works, the issues of the formation, development and evaluation of the effectiveness of cluster-network interactions remain unresolved. The research of the scientific community is based mainly on qualitative methods of cluster analysis (expert analysis, retrospective analysis, comparison method, etc.), however, the need to transform regional development and the transition to neo-economics require the use of economic and mathematical methods of analysis, and their arsenal is relatively small. which necessitates the search for new solutions. An attempt is made in the work to simulate the cluster-network mechanism in the oil and gas industry using neural networks, since the oil sector is one of the key sectors of the Russian economy, which influences the determining rates and paths of the country's socio-economic development, and is subject to the greatest regulation by the government of the country than most other sectors. The most important specific feature of the oil sector is that it is not only able to generate huge monetary resources, but also to accumulate them to solve a large number of socio-economic problems. Based on the results of the trained neural network, using the example of the indicators of the Perm Territory, predicted values of the gross regional product were made and, as a possible core of the oil industry cluster, the profit forecast of the company of the LUKOIL group.