Multistep Traffic Forecasting Using Dynamic Graph Convolution: Interpretations of Real-Time Spatial Correlations
This is the open source code of the proposed Dynamic Graph Convolution (DGC) modules and two corresponding multistep traffic forecasting models: Dynamic Graph Filters Networks (DGFN) and Dynamic Graph GRU (DGGRU) networks.
An example of the dataset used in the article (Rot_cc2) can be downloaded here: https://drive.google.com/file/d/1UCWmA-vLp3LSu1IFSiwdVMXSvdfsVFf9/view?usp=sharing
This is the initial version. A more detailed and proper version will be updated soon.