Predicting Station-Level Short-Term Passenger Flow in a Citywide Metro Network Using Spatiotemporal Graph Convolutional Neural Networks

• Published in: ISPRS international journal of geo-information Vol. 8; no. 6; p. 243
• Main Authors: Han, Yong, Wang, Shukang, Ren, Yibin, Wang, Cheng, Gao, Peng, Chen, Ge
• Format: Journal Article
• Language: English
• Published: Basel MDPI AG 08.06.2019
• Subjects: Airports; Algorithms; Artificial neural networks; Bayesian analysis; citywide metro network; Convolution; deep learning; Engineering; Flow; graph convolution neural network; Inflow; Mathematical models; Model testing; Neural networks; Outflow; Parameter sensitivity; Passengers; Predictions; Principal components analysis; Probability theory; Public safety; Safety management; Spatial dependencies; spatiotemporal flow-volume prediction; Stereograms; Subway stations; Traffic management; Traffic safety; Weekly; China

Predicting the passenger flow of metro networks is of great importance for traffic management and public safety. However, such predictions are very challenging, as passenger flow is affected by complex spatial dependencies (nearby and distant) and temporal dependencies (recent and periodic). In this...