HYBRID ASSIGNED REGRESSIVE AND INTELLECTUAL SYSTEMS FOR FORECASTING RATES OF SOCIAL AND ECONOMIC DEVELOPMENT IN RUSSIA

Abstract

The article deals with general methodology and architecture of the system of hybrid models for forecasting economic indices, its putting into effect in the form of integrated information system illustrated by rates in the field of research and innovation, investment and budget sphere of Russian Federation economy. The authors demonstrate the scheme of work of the assigned information and analytical meta-system, show the general algorithm for the process of retro-verification of typical forecast block, which can raise trust for forecast results. They depict stages of the indices forecast process in the hybrid model. The use of such system could give an opportunity to improve accuracy and quality of forecasts and at the same time to apply them in the contour of managing the attainment of target rates.

Keywords

• сфера исследований и инноваций
• системы регрессионных уравнений
• модели краткосрочного прогноза
• ретроверификация прогноза
• система нейросетевых моделей
• система гибридных моделей