Forecasting the Spread of Viral Diseases in Jordan Using the SARIMA Statistical Model

Abstract

Time series models can help predict disease. Incidence data can be used to predict future disease outbreaks. Advances in modeling techniques allow us to compare the predictive capabilities of different time series models. Public health monitoring systems give essential data for accurate forecasting of future epidemics. This paper describes a study that used two types of infectious disease data, namely Mumps and Chickenpox, collected from a Department of Statistics open source data in mainland Jordan, to assess the performance of time series methods, specifically Seasonal Autoregressive Integrated Moving-Average with Exogenous Regressors. The data collected from 2000 to 2023 were used as modeling and forecasting samples, respectively. The performance was evaluated using two metrics: mean absolute error and mean squared error. The statistical models' accuracy in predicting future epidemic illnesses established their use in epidemiological monitoring. The Seasonal Autoregressive Integrated Moving Average with Exogenous Regressors model, which was used to estimate total mumps cases in Jordan, was applied to a real dataset over the years 2000 to 2023. The dataset was separated into three groups: 78% training, 9% validation, and 13% testing. The results showed a mean squared error of 26629 and a mean absolute error of 152. The model predicted that Jordan will have 2341 cases of mumps by 2028.