Spatio-temporal fusion graph convolutional network for traffic flow forecasting

Ying Ma, Haijie Lou, Ming Yan, Fanghui Sun, Guoqi Li

Abstract: Highlights•Node-specific GCN immensely enhances the capacity of the GCN.•A spatio-temporal fusion module is proposed.•A transformer-based global temporal correlation learning module is proposed.•The results prove that our method is superior to existing competitive methods.