A multi-view bidirectional spatiotemporal graph network for urban traffic flow imputation

Abstract

Accurate estimation of missing traffic data is one of the essential components in intelligent transportation systems (ITS). The non-Euclidean data structure and complex missing traffic flow patterns make it challenging to capture nonlinear spatiotemporal correlations of missing traffic flow, which are critical for the imputation of missing traffic data. In this study, we propose a novel multi-view bidirectional spatiotemporal graph network called Multi-BiSTGN to impute urban traffic data with complex missing patterns. First, three spatiotemporal graph sequences are constructed to comprehensively describe traffic conditions from different temporal correlation views, i.e. temporal closeness view, daily periodicity view, and weekly periodicity view. Then, three bidirectional spatiotemporal graph networks are fused by a parametric-matrix-based method to obtain the final imputation results. To train the Multi-BiSTGN model, a novel loss function that considers the interactions between three temporal correlation views is designed to optimize the parameters of the Multi-BiSTGN model. The proposed model was validated on real-world traffic datasets collected in Wuhan, China. Experimental results showed that Multi-BiSTGN outperformed ten existing baselines under different missing types (random missing, block missing, and mixed missing) and missing rates.

Keywords:

• Graph neural networks
• missing data imputation
• multi-view learning
• spatiotemporal correlations

Acknowledgments

The numerical calculations in this paper have been done on the supercomputing system in the Supercomputing Center of Wuhan University. We are also very grateful to the anonymous reviewers for their suggestions and the editor’s careful revisions.

Data and codes availability statement

The data and codes that support the findings of this study are available in ‘figshare.com’ with the identifier https://doi.org/10.6084/m9.figshare.18489236.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This project was supported by National Key R&D Program of China (International Scientific & Technological Cooperation Program) under grant 2019YFE0106500, National Natural Science Foundation of China under grant 41871308, the Fundamental Research Funds for the Central Universities.

Notes on contributors

Peixiao Wang

Peixiao Wang is a PhD candidate of State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing (LIESMARS), Wuhan University. He received the M.S. degree from The Academy of Digital China, Fuzhou University in 2020. His research focus on spatiotemporal data mining, social computing, and public health.

Tong Zhang

Tong Zhang is a Professor with the State Key Laboratory of Information Engineering in Surveying, Mapping and Remote Sensing (LIESMARS), Wuhan University. He received the M.Eng. degree in cartography and geographic information system (GIS) from Wuhan University, Wuhan, China, in 2003, and the Ph.D. degree in geography from San Diego State University, San Diego, CA, USA, and the University of California at Santa Barbara, Santa Barbara, CA, in 2007. His research topics include urban computing and machine learning.

Yueming Zheng

Yueming Zheng is currently pursuing the M.S. degree with the Aerospace Information Research Institute, Chinese Academy of Sciences. Her research focus on spatiotemporal analysis, spatiotemporal modeling, and remote sensing of environment.

Tao Hu

Tao Hu is an Assistant Professor in Department of Geography at Oklahoma State University. Before joining OSU, he worked as a postdoc research fellow in the Center for Geographic Analysis at Harvard University and the Department of Geography at Kent State University. His research interests include geospatial big data analysis (i.e. social media), health geography, human mobility, and crime geography.