Estimation of COVID-19 cases in South American countries using ARIMA models

Abstract

The main objective of this work is to use ARIMA models for the estimation of new contagions using public data
available for Venezuela and the South American region, currently the main focus of a second COVID19 outbreak. A
30-day prediction is made for the number of Covid-19 cases in South American countries using available public
data. ARIMA models were used to estimate the impact of new contagions on infection dynamics for South America Since the appearance of the first case of the new Covid-19 pneumonia in China, which has become a global public health problem and the great challenge that the infection has represented for the countries of South America to June 24, 2020, a total of 1,866,090 cases have been detected and in the particular case of Venezuela a total of 4,365 cases have been detected for the same date. The rapid increase in the number of cases and the high rate of contagion associated with the virus have led to the development of different mathematical approaches, such as: SIR, SEIR models, neural networks and linear regressions that allow predicting the probable evolution of the epidemic. The ARIMA model has been successfully used in other infections such as influenza, malaria, SARS, among others. In the following work, the 30-day prediction of the number of Covid-19 cases in South American countries is made using public data available. The results of the estimates made using these models show that even in the region, greater efforts are needed to control the epidemic.