FORECASTING MODELS OF INFECTIOUS DISEASES: A REVIEW STUDY

Abstract

The Covid-19 outbreak, which started in Wuhan, China, in December 2019, has spread rapidly worldwide. The unpreparedness of countries' health systems and relevant institutions made the process unmanageable for a certain period. This situation has also affected the activities in scientific fields and has led researchers in different fields to modelling studies for the prediction of infectious diseases. These scientific publications containing predictions were mainly obtained from medical and natural science journals. This study was created to cover mostly Covid-19 studies in the literature, and also included other epidemic prediction studies between 2000 and 2022. Forecasting models are categorised into four groups: SIR model group, logistic model group, time series analysis and machine learning. In addition to these four groups, each has its sub-categories. It also aims to provide information about the studies by describing the models commonly used to estimate the propagation mechanism. SIR model group can be shaped according to infectious diseases, logistic models are similar to data structures of infectious diseases, time series analysis uses lagged values of the variable and Prophet model adds trend, seasonal and holidays to the model. As a result, the strengths and weaknesses of these models are identified according to their characteristics.