Predicting COVID19 pandemic waves including vaccination data with deep learning

Authors: Begga, A., Garibo I Orts, O., De Maria Garcia, S., Escolano, F. , Lozano, M.A. , Oliver, N. , Conejero, J.A.

External link:
Publication: Frontiers of Public health , 2023

We present a deep learning-based approach to predict the number of daily COVID-19 cases in 30 countries, considering the non-pharmaceutical interventions (NPIs) applied in those countries and including vaccination data of the most used vaccines. We empirically validate the proposed approach for four months between January and April 2021, once vaccination was available and applied to the population and the COVID-19 variants were closer to the one considered for developing the vaccines. With the predictions of new cases, we can prescribe NPIs plans that present the best trade-off between the expected number of COVID-19 cases and the social and economic cost of applying such interventions. Whereas mathematical models which include the effect of vaccines in the spread of the SARS-COV-2 pandemic are available, to the best of our knowledge we are the first to propose a data driven method based on recurrent neural networks that considers the waning effect of the immunization acquired either by vaccine administration or by recovering from the illness. This work contributes with an accurate, scalable, data-driven approach to modeling the pandemic curves of cases when vaccination data is available.