METHODOLOGY AND TOOLS FOR PROJECTING FIGURES OF RESEARCH AND INNOVATION SPHERE IN RUSSIA

Abstract

The article deals with general methodology and architecture of the hybrid models system of projecting economic figures and its implementation in the form of the integrated information system illustrated by figures in the sphere of research and innovation in Russian economy. The author describes the sub-system of distributed neuro-network models designed in Python with the help of web-framework Django. The stages in the process of figure projection in the hybrid model were demonstrated. Functional structure of hybrid models underlain by software modules was presented. Such system allows us not only to increase accuracy and quality of projections but to use them in management contour for reaching target indicators.

Keywords

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