Modeling the spread of infections during an epidemiological outbreak using an improved mathematical model

Abstract

Pandemics occur periodically worldwide. An accurate forecasting model is therefore essential to estimate the effect of the pandemic and plan accordingly. This research aims to provide a solution that could help the world predict the number of infection cases during pandemics and prepare to accommodate subsequent cases. The mathematical Multiplicative Holt–Winter (M-HW) model was improved regarding the data used to provide an accurate forecast. The model was applied to the Coronavirus (COVID-19) data, where COVID-19 is the recent pandemic that affected all nations worldwide since 2019. Two different periods in Saudi Arabia were modelled to estimate COVID-19 cases. Based on the daily confirmed cases in February 2023 and February 2022, the model showed accuracy of 99.51 % and 99.66 %, respectively. A MAPE value in February 2023 ranges between 0.015 and 1.07, while it ranges between 0.032 and 2.269 in February 2022. Additionally, the RMSE in February 2023 was 0.35, while in February 2022 it was 6.88. The model proved to be accurate and highly efficient. Thus, M-HW model is useful to forecast the number of cases in different regions in case of a pandemic, which makes a significant contribution to mitigating the spread of the virus minimizing the epidemiological spread impact on healthcare systems and focusing on managing and containing the epidemiological spread.

Keywords

• Pandemic
• Covid-19
• Modelling
• M-HW