As part of this study, the goal was to develop adequate (high-precision) tools that would allow for not only a retrospective, but also a prospective assessment of interregional inequality in living standards in the Russian Federation based on the index of social well-being of the population, which is the result of a convolution of private indices. At first, a hypothesis was put forward about the possibility of building an adequate prognostic (traditional econometric) model of dependence of per capita average monetary incomes of the population on a group of factors. The information base of the study was exclusively official data of regional statistics for 2020–2022. In the course of empirical research (correlation and regression analysis), three econometric models differing in the number of factors (from 2 to 4) were developed. However, they allow (according to the average approximation error, taking values from the interval from 8.8 to 9.6 % for different econometric models) approximating regional statistics data only with an acceptable degree of accuracy. Next, a similar hypothesis was tested, but involving the use of a different tool (index method in combination with artificial intelligence), which makes it possible to measure the dependence of the population’s standard of living on a group of factors. In the course of neuromodelling it was found that any of the 5 artificial neural networks included in the Bayesian ensemble allowed approximating the regional statistics data with a high degree of accuracy (with an average error from 2.8 to 3.9 %). Thus, the second hypothesis can be considered confirmed. As part of the study, the predictive function was implemented by forming a Bayesian ensemble of artificial neural networks. The obtained results of the empirical study can act as a scientific basis for adjusting (updating) the socio-economic policy of regulating the quality and standard of living of the population and its interregional inequality among the constituent entities of the Russian Federation
Keywords
regions of Russia, interregional inequality, standard of living, cash income, index method, correlation-regression analysis, artificial intelligence, forecasting