AUTHOR=Šušteršič Tijana , Blagojević Andjela , Cvetković Danijela , Cvetković Aleksandar , Lorencin Ivan , Šegota Sandi Baressi , Milovanović Dragan , Baskić Dejan , Car Zlatan , Filipović Nenad 

TITLE=Epidemiological Predictive Modeling of COVID-19 Infection: Development, Testing, and Implementation on the Population of the Benelux Union

JOURNAL=Frontiers in Public Health

VOLUME=Volume 9 - 2021

YEAR=2021

URL=https://www.frontiersin.org/journals/public-health/articles/10.3389/fpubh.2021.727274

DOI=10.3389/fpubh.2021.727274

ISSN=2296-2565

ABSTRACT=Since the outbreak of the new coronavirus COVID-19, the whole world has taken interest in the mechanisms of its spread and development. Mathematical models have been valuable instruments for the study of the spread and control of infectious diseases. For that purpose, we propose two-way approach in modelling COVID-19 spread: SEIRD model based on differential equations and LSTM deep learning model. SEIRD model is a compartmental epidemiological model with included components - susceptible, exposed, infected, recovered, deceased. In the case of SEIRD model, official statistical data available online for countries of Benelux (Belgium, Netherlands and Luxembourg) in the period of March 15th 2020 to March 15th, 2021 were used. Based on them, we have calculated the key parameters and forward them to the epidemiological model which will predict the number of infected, deceased and recovered people. Results show that SEIRD model is able to accurately predict several peaks for all three countries of interest, with very small RMSE, except for the mild cases (maximum RMSE was 240.79±90.556), which can be explained by the fact that no official data was available for mild case, but this number was derived from other statistics. On the other hand, LSTM represents a special kind of recurrent neural network structure that can comparatively learn the long-term temporal dependencies. Results show that LSTM is capable to predict several peaks based on the position of previous peaks with the RMSE low values. Higher values of RMSE are observed in the cases of infected cases in Belgium (RMSE was 535.93) and Netherlands (RMSE was 434.28), which is expected due the thousands of infected people per day in those countries. In future studies, we will extend the models to include mobility information, variants of concern, as well as a medical intervention, etc. A prognostic model could help us predict epidemic peaks. In that way, we could react in a timely manner by introducing new or tightening existing measures before the health system is overloaded.