Integrated approach to forecasting of socio-economic development for regions

Abstract

Purpose: the author develops a methodology for multidimensional forecasting of economic processes using a matrix model. The construction of this model is based on the joint use of regression models and the apparatus of main components. Discussion: the apparatus for forecasting regional processes constantly requires improvement. Their multi-dimensionality and complexity of interaction practically exclude the possibility of using traditional methods and models. The author proposed a two-level system for calculating balanced forecast estimates in the developed methodology. This methodology allows taking into account both the overall system balance and the internal block balance. This effect is achieved by using the main components of the second level. The main components of the second level are the main components built on the calculated values of ordinary main components. In accordance with this two-level approach, the author provided the construction of matrix models, which give multidimensional forecast estimates at each level. Results: the author proposed formation method of the multidimensional forward-looking way of systematically balanced assumptions.