Precise prediction of the trend of the pandemic has been crucial in COVID-19, as it allows the governments and healthcare systems to effectively distribute and prioritize resource allocations. In this issue, Jo et al. present a neural network model to make quantile predictions on COVID-19 death tolls at the county level. This fine geographical scale, which is rare in publicly available models, can benefit state-level officials to coordinate resources within the state. The quantile predictions also inform people about the distribution of the predicted death tolls, allowing better evaluation of possible trajectories of the severity. The model’s scalability and generalizability further suggests its application in generating other useful predictions such as new cases or hospitalizations by incorporating additional data sources. For details please refer to the article by Jo et al. in pp. 125‒138.
