Análise da aplicação do modelo ARIMA: estudo em uma instituição federal de ensino superior

Abstract

Stocks planning is a challenging task for managers of both private and public sectors. In order to provide a better quality of public spending and transparency, public managers have been adopting scientific techniques instead of empirical practices. In this sense, the present study sought to analyze and apply the ARIMA method in the forecast of internal demand (consumption) from historical data in to Warehouse of a Federal Institution of Higher Education. Specifically, it was contextualized the procedures adopted by the institution, determined which products should be studied and historical consumption data were collected. Thus, based on an exploratory-descriptive case study, a demand forecast was elaborated using the ARIMA and Box-Jenkins Methods. The ARIMA(0,1,2) model was the best possible for a series, according to the AIC and BIC criteria and residue analysis. Therefore, was verified that the model found can be used, but needs attention as to its accuracy.