Long-term forecasting of decomposition time series of socio- economic indicators

Abstract

Purpose: the article presents the results of work on the creation of an adapted complex methodology for predicting the dynamics of decompositional time series of tourist flow in the ski village of Dombay, the features of which are the joint use of both classical and new «nonlinear» statistics. Discussion: the proposed and tested methods are presented in the form of a pre-forecast and forecast model for assessing the trend stability of time series of tourist flow and obtaining a forecast. The following methods of nonlinear dynamics have been tested: the Hearst normalized span method, phase analysis, and linear cellular automaton. Results: the results of the analysis and forecast on real data of the tourist flow are presented in the form of values of the lower level of modeling of tourist and recreational activities, which in turn are input data for models of the upper level – the level of management of tourist and recreational activities. The quantitative forecast of the size of the tourist flow allows us to solve the issues of managing tourist and recreational activities, for example, in planning the employment of the number of rooms.