Authors: Derick Welman, Uriel Calvi
With the first cases reported in December 2019 in Wuhanna China, the SARS-CoV-2 (COVID-19) virus pandemic was officially recognized by the World Health Organization (WHO) on March 11, 2020. At the last week of April of the same year, Brazil, with a total of 3670 deaths, already occupied the 11th place in the ranking of countries with more deaths. Despite the scenario, in Brazil the restrictive measures were and still are a matter of debate and the federal government's decision on the subject varies until today. Considering the contribution of forecasting tools in decision-making situations, the objective of this work is to evaluate the performance of the ARIMA and Neural Networks models in predicting the number of deaths through a COVID-19 time series. In conclusion, we observed that the model that best adapted to the data was ARIMA Seasonal, but the data can still be improved to improve the training of the models.