Synthesis of nonlinear dynamics methods and regression analysis for the study of socio-economic processes

In this article the synthesis of regression analysis methods with the tools of nonlinear dynamics analysis is proposed. Such an approach proves effective in case of analysis of empirical data of nonlinear processes developing under conditions of constraint, provided that the increments of processes are proportional to the achieved level adjusted for the constraints. Application of the method proposed is demonstrated by example of the analysis of revenue dynamics of Russian large and medium-sized companies from 2009 to 2015. The generalizing results obtained allow to classify the companies by the nature of changes in revenue as well as to assess their current positions and further dynamics. The models are proposed for different mechanisms of process developing which manifest in a variety of constraints.