Modeling epidemic growth curves using nonlinear rational polynomial equations: an application to Brazil's COVID-19 data

Ricardo Puziol De Oliveira, Jorge Alberto Achcar, Wesley Bertoli, Josmar Mazucheli, Yara Campos Miranda


This paper reports a broad study using epidemic-related counting data of COVID-19 disease caused by the novel coronavirus (SARS-CoV-2). The considered dataset refers to Brazil's daily and accumulated counts of reported cases and deaths in a fixed period (from January 22 to June 16, 2020). For the data analysis, it has been adopted a nonlinear rational polynomial function to model the mentioned counts assuming Gaussian errors. The least-squares method was applied to fit the proposed model. We have noticed that the curves are still increasing after June 16, with no evidence of peak being reached or decreasing behavior in the period for new reported cases and confirmed deaths by the disease. The obtained results are consistent and highlight the adopted model's capability to accurately predict the behavior of Brazil's COVID-19 growth curve in the observed time-frame.


COVID-19 counting data; Gaussian errors; Nonlinear models; Rational polynomial functions; SARS-CoV-2

Texto completo:

PDF (English)

DOI: 10.3895/rts.v18n50.13534


  • Não há apontamentos.

Direitos autorais 2021 CC-BY

Licença Creative Commons
Esta obra está licenciada sob uma licença Creative Commons Atribuição 4.0 Internacional.

Av. Sete de Setembro, 3165 - Rebouças CEP 80230-901 - Curitiba - PR - Brasil