Spatial Temporal Problem by Graph Model

[ Collaborated with DAC at Virginia Tech ]

Graph neural networks (GNNs) motivate many applications based on network data such as transportation road. However, as a practical scenario, the road has very different characteristics from a smooth network (e.g., social network) often used by GNN experiments. We plan to identify these special features and accordingly adjust GNN to adopt transportation problems.