Forecasting of buffalo milk in a Brazilian diary using the ARIMA model

Abstract

Brazil presents a great diversity in herd rearing. These herds are divided into nine categories that are of great importance, both for domestic consumption and for export. Among these nine groups, buffaloes are the type of herd that has the least amount of animals in breeding stock. Despite presenting a small number of heads, the consumption of milk and dairy products in this group has been growing in recent years. In view of this situation, the objective of this article is to use the ARIMA time series model to model and forecast the milk demand of a buffalo dairy for one year and thus align product demand with production demand and inventory. For the analysis and construction of the model, a time series from a Brazilian dairy was used. These data represent the demand for raw materials from January 2011 to December 2016. The model was based on historical dairy data only. Initially, it is observed that the historical series shows trend and seasonality. After the data modelling, the forecast model reached was the ARIMA model (2,1,4) (0,1,1)12, through Minitab software.