The use of series study for forecasting is particularly relevant. Its use in a business context can somehow adjust the business strategy. In this context, the work reported in this article corresponds to an empirical study in which the best way to predict using several classical methods in the context of econometrics was identified. The results presented show significant differences in function of the analysed realities. Forecast errors associated with incidents are less than orders or invoicing. One of the relevant results of the work presented here was the creation of a solution that can be used by managers for decision making.
Tomás, Daniel; Costa, Carlos J.; Gaivão, José Pedro; and Carvalho, João Paulo, "Time series for incidences, orders and invoicing forecast" (2018). 2018 Proceedings. 36.