Nonlinear regression in COVID-19 forecasting

ABSTRACT

This paper introduces some kinds of nonlinear growth curves for forecasting cumulative COVID-19 patients. It is shown that the Richards curve is reasonable and flexible in this COVID-19 forecasting. The nonlinear growth curve regression model is established for forecasting cumulative COVID-19 patients and the parameter estimation approach for the model is also given. Specifically, the COVID-19 situation forecasting in China is made well which includes forecasting based on consecutive and piecewise time fitting. It provides a good basis for the future work. © 2021, Science China Press. All right reserved.