Simulation and application of COVID-19 compartment model using physic-informed neural network

In this work, SVEIDR model and its variants (Aged, Vaccination-structured models) are introduced to encode the effect of social contact for different age groups and vaccination status. Then we implement the Physic-Informed Neural Network on both simulation and real-world data. Results including the spread and forecasting analysis of COVID-19 learned from the neural network are shown in the paper.