Graph enhanced spatial-temporal transformer for traffic flow forecasting

Weishan Kong, Yanni Ju, Shiyuan Zhang, Jun Wang, Liwei Huang, Hong Qu

Abstract: Highlights•We propose an effective GE-STT model for traffic flow forecasting.•A GE module is designed to capture the intricate spatial and temporal features.•We introduce a correction term to rectify potential errors.•Extensive experiments show the superior performance of GE-STT.