RuEn

Journal section "Theoretical and methodological issues"

Forecasting the Indicators of Scientific, Technological and Innovative Development of the Region Using Recurrent Neural Networks

Byvshev V.I., Koroleva S.A., Panteleeva I.A., Pisarev I.V.

Volume 17, Issue 3, 2024

Byvshev V.I., Koroleva S.A., Panteleeva I.A., Pisarev I.V. (2024). Forecasting indicators of scientific, technological and innovative development of the region using recurrent neural networks. Economic and Social Changes: Facts, Trends, Forecast, 17(3), 102–117. DOI: 10.15838/esc.2024.3.93.6

DOI: 10.15838/esc.2024.3.93.6

Abstract   |   Authors   |   References
The article forecasts indicators of scientific, technological and innovative development of a constituent entity of the Russian Federation and regional institutions of innovative development using recurrent neural networks. Forecasting using neural networks has become widespread and is a relevant, high-quality and reliable way of making economic forecasts and is applicable within the framework of socio-economic analysis, including analysis of territories. However, when studying the scientific literature, it was not possible to find works in which the scientific, technological and innovative development of regions was predicted using the neural network method, which determines the scientific novelty of the research being carried out. The relevance of the study is due to the increasing attention on the part of regional authorities to the scientific, technological and innovative development of territories and the need to form state programs of the constituent entities of the Russian Federation in the field of scientific and technological development. The research hypothesis is that forecasting indicators of scientific, technological and innovative development of the region and the activities of regional institutions for innovative development using recurrent neural networks will give more accurate results than using the linear regression method, moving average model or the Holt – Winters method. As part of the study, a recurrent neural network model was formed based on a system of interconnection of indicators of scientific, technological and innovative development of a constituent entity of the Russian Federation and regional institutions of innovative development. As a result, a forecast of indicators of scientific, technological and innovative development of a constituent entity of the Russian Federation and the activities of regional institutions for innovative development was obtained, which correlates with the real situation in this area

Keywords

regional scientific and technological policy, innovative development institutions, recurrent neural networks, forecasting, scientific and technological development indicators, regional economy

View full article