SPLINEMODELS IN FORECASTING THE BALANCE OF DEVELOPMENT OF THE INDUSTRIAL COMPLEX

Abstract

The article covers the possibility and necessity of using splines for analyzing the current dynamics and forecasting stochastic, turbulent and poorly predictable macroeconomic indicators characterizing the balance of industrial development. The essence, advantages, and types of spline modeling as the main research method are analyzed. The expediency of using the spline or spline-polynomial approach in modeling and forecasting the macroeconomic dynamics of the studied balance indicators is confirmed by the formation of a model consisting of successively connected fragments of its change. At the beginning of the study, the selection of the type of spline was made on the basis of a graphic representation of the dynamics of the index of entrepreneurial confidence related to the level of balance in the development of industrial enterprises. It is concluded that correlation splines are most acceptable. Graphic analysis of averaged data of changes in the index of entrepreneurial confidence over the years 2005-2017 allowed to identify 5 fragments describing the annual cycle of changes in the studied indicator. Based on this, the concept of continuous spline approximation of entrepreneurial confidence in the form of a spline-representation of the macroeconomic dynamics of this indicator on the initial discrete set of observation points during the analyzed period has been implemented. The article shows the simplest use of the spline paradigm, which can be extended by a spline approximation of the change in the average value of the index of entrepreneurial confidence during the period studied (2005-2017) in the form of a periodic process