Predictive models for the number of cumulative cases for spreading coronavirus disease 2019 in the world

Abstract

The coronavirus disease outbreak in 2019 (COVID-19) has caused major economic and healthcare problems worldwide. At this time, the worldwide outbreak has passed its peak, while the greatest number of cases has been in the USA, Brazil, and India. Measures and policies for controlling the outbreak have been developed by authorities to protect the population of each country, and forecasting the number of infectious people is an important factor for developing them. This research was conducted to identify a suitable forecasting model for estimating the cumulative daily number of infectious people worldwide. Sample countries with severe outbreaks were selected from each continent. Herein, forecasting models based on logistic, Richards, and Gompertz growth curves are derived and their suitability for forecasting the COVID-19 rates in each sample country and worldwide are analyzed. Moreover, estimating the growth curve parameters is based on the least-squares method. The results show that the Gompert growth curve is the most suitable for estimating the cumulative number of infectious people worldwide.

Keywords

• coronavirus disease 2019
• logistic growth curve
• richards growth curve
• gompertz growthcurve
• least square estimation