Author(s): Claudimar Pereira Da Veiga, Cássia Rita Pereira Da Veiga, Anderson Catapan, Ubiratã Tortato, Wesley Vieira Da Silva
Abstract: For business operations in retail companies which work with food products with short life cycle and perishables, the accuracy of forecast is of crucial importance because of the volatile demand pattern, influenced by an environment of rapid and dynamic response. In several studies in the literature, the choice of the most suitable forecasting model remains a central concern. In this context, this article aims to compare the performances between ARIMA and Holt-Winters (HW) models for the prediction of a time series formed by a group of perishable dairy products. As performance measures, metric analysis of the Mean Absolute Percentage Error (MAPE) and the Theil inequality index (U-Theil) were used. In this study, the HW model obtained better results regarding the performance metrics, having a better adjustment and capturing the linear behavior of the series.