Aplicação da modelagem matemática para descrição de casos cumulativos de covid-19 em Uruçuí - PI

Abstract

The internalization of confirmed cumulative cases of COVID-19 in the city of Uruçuí-PI raises questions about the development of virus contamination in the city. Thus, the municipality's Secretary of Health provided the number of cases per day in the period from May 1, 2020 to December 12, 2020, so that linear and non-linear, Logistic and Gompertz models, adjusted adjusted. For residual analysis, the Shapiro-Wilk tests were used to verify normality, Breusch Pagan (Zeileis, Hothorn, 2002) for variance homogeneity and Durbin-Watson (Zeileis, Hothorn, 2002) for independence. Model parameters were estimated by the least squares method using the Gauss-Newton algorithm. The corrected Akaike information criteria and the adjusted determination coefficient were used as adjustment meters to choose the best model. Among the adjusted models, it is concluded that the Gompertz model is suitable for describing the cumulative cases of COVID-19 in Uruçuí-PI. Analyzes were performed using the R software. It is noteworthy that the adjusted model cannot be used for future forecasts, as it was tested only for the period studied. However, the result of these models, which were mainly adjusted to observe the spread of the virus, directly helps the authorities in controlling the spread of the disease and preventing measures that must be taken.