Прогнозування об'ємів продажу магазинів взуття

Abstract

The article presents the idea of developing software for forecasting sales volumes of shoe stores based on additive and multiplicative time series models. The program should make it possible to analyze the impact of seasonal components on sales volumes for a certain quarter, obtain model estimates and build a forecast of sales olumes for several future quarters based on the best model. As a result of the work, the input time series were smoothed using the moving average method and the seasonal component was estimated. The seasonal component was removed from the levels of the series and smoothed data were obtained. The parameters of the trend line equation were found using the method of least squares. The adequacy of the models was also checked, and on the basis of the best one, a forecast of sales volumes for the next few quarters was made.