Incorporating variant frequencies data into short-term forecasting for COVID-19 cases and deaths in the USA: a deep learning approach

This repository provides the code and data for the Multi-stage LSTM model to Forecast Short-Term COVID-19 Cases and Deaths in the US!

The paper can be found here: (https://www.thelancet.com/journals/ebiom/article/PIIS2352-3964(23)00047-6/fulltext).

A) Network architecture of the multi-stage LSTM model. B) Prediction structure of the multi-stage LSTM model. At the initial stage, the model uses the most recent data as input, then at the later stage, the model adapts previous prediction as input to make further predictions. The transparent colors represent the model’s output, and solid colors represents the model’s inputs. C) An example forecasting of the multi-stage LSTM model.

Data

State-Level Data	Data Preprocessing	Data Source
COVID-19 cases/deaths	Raw	https://github.com/CSSEGISandData/COVID-19
Growth rate of cases/deaths	Raw	https://github.com/CSSEGISandData/COVID-19
Vaccination coverage	Raw	https://www.cdc.gov/nhsn/covid19/dial-vaccination-dashboard.html
Hospitalization data	Raw	https://www.hhs.gov/index.html
Importation risk	Derived	https://www.safegraph.com/, https://github.com/CSSEGISandData/COVID-19
Mobility ratio	Derived	https://www.safegraph.com/
Visits ratio for 21 different destinations	Derived	https://www.safegraph.com/
COVID-like symptoms in community	Raw	https://github.com/cmu-delphi/delphi-epidata
Temperature & Percipitation	Raw	https://github.com/CSSEGISandData/COVID-19_Unified-Dataset
Demographic data	Raw	https://www.census.gov/data.html
Variant cases	Derived	https://gisaid.org/

Please refer to the Appendix for detailed preprocessing for derived metrics.

Model

• Main model: Predict the target epidemiological variables of interest!
• Feature model: Predict independent features to populate the data streams used as input in the main model!
• Prediction: The multi-stage model builds off the initial first stage prediction to forecast an additional week out and continues to implement this iterative approach one stage at a time, to predict further into the future.

Contributors

Hongru Du, Ensheng Dong, Hamada S. Badr, Mary E. Petrone, Nathan D. Grubaugh, and Lauren M. Gardner

BibTeX

@article{du2023incorporating,
  title={Incorporating variant frequencies data into short-term forecasting for COVID-19 cases and deaths in the USA: a deep learning approach},
  author={Du, Hongru and Dong, Ensheng and Badr, Hamada S and Petrone, Mary E and Grubaugh, Nathan D and Gardner, Lauren M},
  journal={EBioMedicine},
  volume={89},
  year={2023},
  publisher={Elsevier}
}