Integrated system analysis and forecasting of non-stationary processes

Abstract

This article proposes an integrated system analysis and forecasting of non-stationary time series in order to improve the adequacy of the developed models of non-stationary processes. The methodology is proposed for modeling non-stationary processes that includes the following stages: the preliminary data processing, estimation of the model structure and its parameters, computing of forecasts estimates. Examples are provided of the use of this system for the analysis and prediction of prices of Food Production Company and for the financial processes prediction. The results indicate that the integrated system for the analysis, modeling, and forecasting of non-stationary processes performs automated processing of data, automatically determines the structure of the model, and selects the best forecasting models.