Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction

The daily long-term traffic prediction is an important urban computing issue, and can give users a global insight into traffic. Accurate traffic prediction is conducive to rational route planning and efficient traffic resource allocation. However, it is challenging to capture the global spatial-temp...

Full description

Saved in:
Bibliographic Details
Main Authors: Song Zhang, Yanbing Liu, Yunpeng Xiao, Rui He
Format: Article
Language:English
Published: Springer 2022-11-01
Series:Journal of King Saud University: Computer and Information Sciences
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S1319157822002993
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1849325025089290240
author Song Zhang
Yanbing Liu
Yunpeng Xiao
Rui He
author_facet Song Zhang
Yanbing Liu
Yunpeng Xiao
Rui He
author_sort Song Zhang
collection DOAJ
description The daily long-term traffic prediction is an important urban computing issue, and can give users a global insight into traffic. Accurate traffic prediction is conducive to rational route planning and efficient traffic resource allocation. However, it is challenging to capture the global spatial-temporal correlations for daily long-term traffic prediction. In this paper, we propose a spatial-temporal upsampling graph convolutional network (STUGCN) for daily long-term traffic speed prediction. STUGCN uses an innovative upsampling method to capture the global spatial-temporal correlations. Specifically, in spatial dimension, we construct an upsampled road network by adding virtual nodes to the original road network to capture local and global spatial correlations. In temporal dimension, we build a time graph to capture the temporal correlations among adjacent time steps. Besides, we construct a knowledge base, and the global temporal correlations can be captured by upsampling the current day from the knowledge base. Therefore, STUGCN not only preserves the local spatial-temporal correlations, but also has the ability to learn global spatial-temporal correlations. The experimental results on two real-world datasets demonstrate that our approach is approximately 16.4%–17.1%, 14.1%–17.0% and 17.4%–22.4% better than the state-of-the-art in terms of MAE, RMSE and MAPE metrics, respectively.
format Article
id doaj-art-8566d276ed3845bd8f14996c7fb43900
institution Kabale University
issn 1319-1578
language English
publishDate 2022-11-01
publisher Springer
record_format Article
series Journal of King Saud University: Computer and Information Sciences
spelling doaj-art-8566d276ed3845bd8f14996c7fb439002025-08-20T03:48:31ZengSpringerJournal of King Saud University: Computer and Information Sciences1319-15782022-11-0134108996901010.1016/j.jksuci.2022.08.025Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed predictionSong Zhang0Yanbing Liu1Yunpeng Xiao2Rui He3School of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaSchool of Medical Information, Chongqing Medical University, Chongqing 400065, China; Corresponding author.School of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaSchool of Computer Science and Technology, Chongqing University of Posts and Telecommunications, Chongqing 400065, ChinaThe daily long-term traffic prediction is an important urban computing issue, and can give users a global insight into traffic. Accurate traffic prediction is conducive to rational route planning and efficient traffic resource allocation. However, it is challenging to capture the global spatial-temporal correlations for daily long-term traffic prediction. In this paper, we propose a spatial-temporal upsampling graph convolutional network (STUGCN) for daily long-term traffic speed prediction. STUGCN uses an innovative upsampling method to capture the global spatial-temporal correlations. Specifically, in spatial dimension, we construct an upsampled road network by adding virtual nodes to the original road network to capture local and global spatial correlations. In temporal dimension, we build a time graph to capture the temporal correlations among adjacent time steps. Besides, we construct a knowledge base, and the global temporal correlations can be captured by upsampling the current day from the knowledge base. Therefore, STUGCN not only preserves the local spatial-temporal correlations, but also has the ability to learn global spatial-temporal correlations. The experimental results on two real-world datasets demonstrate that our approach is approximately 16.4%–17.1%, 14.1%–17.0% and 17.4%–22.4% better than the state-of-the-art in terms of MAE, RMSE and MAPE metrics, respectively.http://www.sciencedirect.com/science/article/pii/S1319157822002993Long-term traffic speed predictionSpatial-temporal upsamplingGraph convolutional networkIntelligent transportation system
spellingShingle Song Zhang
Yanbing Liu
Yunpeng Xiao
Rui He
Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
Journal of King Saud University: Computer and Information Sciences
Long-term traffic speed prediction
Spatial-temporal upsampling
Graph convolutional network
Intelligent transportation system
title Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
title_full Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
title_fullStr Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
title_full_unstemmed Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
title_short Spatial-temporal upsampling graph convolutional network for daily long-term traffic speed prediction
title_sort spatial temporal upsampling graph convolutional network for daily long term traffic speed prediction
topic Long-term traffic speed prediction
Spatial-temporal upsampling
Graph convolutional network
Intelligent transportation system
url http://www.sciencedirect.com/science/article/pii/S1319157822002993
work_keys_str_mv AT songzhang spatialtemporalupsamplinggraphconvolutionalnetworkfordailylongtermtrafficspeedprediction
AT yanbingliu spatialtemporalupsamplinggraphconvolutionalnetworkfordailylongtermtrafficspeedprediction
AT yunpengxiao spatialtemporalupsamplinggraphconvolutionalnetworkfordailylongtermtrafficspeedprediction
AT ruihe spatialtemporalupsamplinggraphconvolutionalnetworkfordailylongtermtrafficspeedprediction