This forecasting model combines the signal of a discrete stochastic epidemic computational model GLEAM (Global Epidemic and Mobility Model) with a deep learning spatiotemporal forecasting framework to further improve predictions to study the spatiotemporal COVID-19 spread. The hybrid model leverages rich real-world data about COVID-19: when a person has been infected; where they have traveled; death records; travel and re-opening restrictions from all 50 states. DeepGLEAM is a part of the national ensemble forecast at the CDC (Centers for Disease Control). Projections will be regularly updated as new data and information about mitigation policies become available. Sensitivity analysis on the basic parameters is routinely performed along with the baseline projections considered. In order to calculate the number of deaths the model uses estimates of COVID-19 severity from available data [1, 2].
Dongxia (Allen) Wu, Liyao Gao, Xinyue Xiong, Matteo Chinazzi, Alessandro Vespignani, Yian Ma, Rose Yu
University of California, San Diego, Northeastern University
Disclaimer: There are large uncertainties around the transmission of COVID-19, the effectiveness of different policies and the extent to which the population is compliant to social distancing measures. The presented material is based on modeling scenario assumptions informed by current knowledge of the disease and subject to change as more data become available. Future decisions on when and for how long to relax mitigation policies will be informed by ongoing surveillance. Additional modeling and data studies are required to assess the level and effectiveness of additional non-pharmaceuticals interventions required to lift current social distancing measures.