Application of trend-seasonal models for time series’ research and forecasting

Abstract

Purpose: the authors consider the application of trend-seasonal models for the analysis of time series and their forecasting. The authors suggest using short-term forecasting models that have established themselves as high- precision tools when working with trend-seasonal time series. Discussion: the authors analyzed two temporary sets of sales volumes according to the objectives of the study. It is worth noting that the authors applied the additive and multiplicative forecasting models for their study due to the presence of a seasonal component in the time series under consideration. Results: the analysis of time series made it possible to forecast sales dynamics, which allows to identify the trends in their decline and growth, as well as to compare the forecast estimates of the applied models and the source data. The authors proposed a methodology for the formation of short-term forecasts, the basis for which was the mathematical apparatus of the trend-seasonal modeling, using the built in functions of the MS Excel spreadsheet processor and the graphical visualization tools available in it.