Modelling spreading of an infection using time series by a novel family of models; fitting the time series of the confirmed cases of COVID-19 in China

Abstract

Abstract Introduction Time series models are one of the frequently used methods to describe the pattern of spreading an epidemic. Methods We presented a new family of time series models able to represent the cumulative number of individuals that contracted an infectious disease from the start to the end of the first wave of spreading. This family is flexible enough to model the propagation of almost all infectious diseases. After a general discussion on competent time series to model the outbreak of a communicable disease, we introduced the new family through one of its examples. Results We estimated the parameters of two samples of the novel family to model the spreading of COVID-19 in China. Discussion Our model does not work well when the decreasing trend of the rate of growth is absent because it is the main presumption of the model. In addition, since the information on the initial days is of the utmost importance for this model, one of the challenges about this model is modifying it to get qualified to model datasets that lack the information on the first days.