Spatial-Temporal Multi-Granularity Framework for Accurate and Robust Traffic Forecasting
A novel multi-granularity framework that enhances the capture of long-distance and long-term dependencies in traffic networks, enabling accurate and robust traffic forecasting.