МЕТОДИЧЕСКИЙ ПОДХОД К ПРОГНОЗИРОВАНИЮ ЦЕЛЕВЫХ ПАРАМЕТРОВ СОЦИАЛЬНО-ЭКОНОМИЧЕСКОГО РАЗВИТИЯ РЕГИОНОВ С ИСПОЛЬЗОВАНИЕМ СТАНДАРТНОГО АППАРАТА НЕЙРОННЫХ СЕТЕЙ

Abstract

S u b j e ct . Methods of forecasting target parameters of socio - economic development of regions. R e s e a r c h topic. Forecasting the export of Russian regions using the standard apparatus of neural networks. P u r p o s e. Development and testing of a methodological approach to forecasting the target parameters of the socio - economic development of Russian regions for the period up to 2024. M e t h o d o l o g y. Logical, comparative, economic - statistical, cluster analysis, neural network prediction. R e s u l t s . Five virtual clusters (groups of regions) are formed, with a sufficient level of reliability and significance of parameters. An analysis of the export dynamics of model regions representing different groups (virtual clavesters) of Russian regions, differing in the totality of parameters of foreign economic activity, was made. A forecast of the dynamics of exports of model regions has been formed. S c o p e . Economic policy of regional authorities in the field of foreign economic activity. C o n c l u s i o n s . During the study, it was found that for most region groups, the combination of cluster analysis, correlation - regression and neural network prediction methods allows: to establish the composition of basic indicators of socio - economic development associated with the target parameter; make at least two variants of target parameter dynamics forecast; to determine the degree of difference between the forecast variants of the target parameter dynamics; determine the acceptability of using formal prediction methods (correlation - regression and neural) to predict the dynamics of the target parameter of socio - economic development of model regions.