A comparative study to choose the appropriate growth model to forecast COVID-19 cases in Iraq

Abstract

COVID-19 infection cases forecasting is a process of estimating future values based on historical data which is playing an important role in health decision making in various fields. Daily infection cases of COVID-19 can be considered as a time series represent the growth of the number of infected persons in a population. Consequently, the growth models may be used to forecast any population growth such as population of infected people with the Covid-19 virus. The popular models of growth such as logistic, log-logistic, Gompertz, Weibull and Richards models are useful to describe the growth of many phenomena like an epidemic and the spread of the number of infected people. The main objective of this paper is to choose a successful growth model after comparing these models to make good use of the current data on COVID-19 in Iraq to better understand the spread of this disease and to forecast the future daily infection cases. AIC, BIC and other goodness of fit criteria and daily infection cases in Iraq for the period from 1st Jan. 2021 until 30th April 2021 were used to compare these models and choose the successful model. The results of fitting these model show that the appropriate models are Weibull type 1 and log-logistic with five parameters models, and the predicted numbers of infected cases are near the actual numbers of infected cases.