Group-Aware Neural Network

What is GAGNN? GAGNN, or Group-aware Graph Neural Network, is a powerful model for nationwide city air quality forecasting. It is designed to construct a city graph and a city group graph to model the spatial and latent dependencies between cities in order to forecast air quality. By introducing a differentiable grouping network to identify the latent dependencies among cities and generate city groups, GAGNN can more effectively capture the dependencies between city groups. How Does GAGNN Wor