WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting

Regional traffic forecasting is a critical challenge in urban mobility, with applications to various fields such as the Internet of Everything. In recent years, spatio-temporal graph neural networks have achieved state-of-the-art results in the context of numerous traffic forecasting challenges. Thi...

Full description

Saved in:
Bibliographic Details
Main Authors: Theodoros Theodoropoulos, Angelos-Christos Maroudis, Uwe Zdun, Antonios Makris, Konstantinos Tserpes
Format: Article
Language:English
Published: Elsevier 2025-06-01
Series:International Journal of Information Management Data Insights
Subjects:
Online Access:http://www.sciencedirect.com/science/article/pii/S2667096825000205
Tags: Add Tag
No Tags, Be the first to tag this record!
_version_ 1850225638519078912
author Theodoros Theodoropoulos
Angelos-Christos Maroudis
Uwe Zdun
Antonios Makris
Konstantinos Tserpes
author_facet Theodoros Theodoropoulos
Angelos-Christos Maroudis
Uwe Zdun
Antonios Makris
Konstantinos Tserpes
author_sort Theodoros Theodoropoulos
collection DOAJ
description Regional traffic forecasting is a critical challenge in urban mobility, with applications to various fields such as the Internet of Everything. In recent years, spatio-temporal graph neural networks have achieved state-of-the-art results in the context of numerous traffic forecasting challenges. This work aims to expand upon the conventional spatio-temporal graph neural network architectures in a manner that may facilitate the inclusion of information regarding the examined regions and the populations that traverse them to establish a more efficient prediction model. The end-product of this scientific endeavor is a novel spatio-temporal graph neural network architecture for regional traffic forecasting referred to as WEST (WEighted STacked) GCN-LSTM. Furthermore, the aforementioned information is included via two novel dedicated algorithms, the Shared Borders Policy and the Adjustable Hops Policy. Through information fusion and distillation, the proposed solution significantly outperforms its competitors in an experimental evaluation of 19 forecasting models across several datasets. Finally, an additional ablation study determined that each component of the proposed solution enhances its overall performance.
format Article
id doaj-art-eb2fbeb123f343d6bbbd40b13dd3a117
institution OA Journals
issn 2667-0968
language English
publishDate 2025-06-01
publisher Elsevier
record_format Article
series International Journal of Information Management Data Insights
spelling doaj-art-eb2fbeb123f343d6bbbd40b13dd3a1172025-08-20T02:05:17ZengElsevierInternational Journal of Information Management Data Insights2667-09682025-06-015110033810.1016/j.jjimei.2025.100338WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecastingTheodoros Theodoropoulos0Angelos-Christos Maroudis1Uwe Zdun2Antonios Makris3Konstantinos Tserpes4Harokopio University of Athens, Omirou 9, Athens, 17778, Greece; University of Vienna, Software Architecture Research Group, Vienna, 1090, Austria; Corresponding author at: Harokopio University of Athens, Omirou 9, Athens, 17778, Greece.National Technical University of Athens, Heroon Polytechniou 9, Athens, 15780, GreeceUniversity of Vienna, Software Architecture Research Group, Vienna, 1090, AustriaNational Technical University of Athens, Heroon Polytechniou 9, Athens, 15780, Greece; Harokopio University of Athens, Omirou 9, Athens, 17778, GreeceNational Technical University of Athens, Heroon Polytechniou 9, Athens, 15780, Greece; Harokopio University of Athens, Omirou 9, Athens, 17778, GreeceRegional traffic forecasting is a critical challenge in urban mobility, with applications to various fields such as the Internet of Everything. In recent years, spatio-temporal graph neural networks have achieved state-of-the-art results in the context of numerous traffic forecasting challenges. This work aims to expand upon the conventional spatio-temporal graph neural network architectures in a manner that may facilitate the inclusion of information regarding the examined regions and the populations that traverse them to establish a more efficient prediction model. The end-product of this scientific endeavor is a novel spatio-temporal graph neural network architecture for regional traffic forecasting referred to as WEST (WEighted STacked) GCN-LSTM. Furthermore, the aforementioned information is included via two novel dedicated algorithms, the Shared Borders Policy and the Adjustable Hops Policy. Through information fusion and distillation, the proposed solution significantly outperforms its competitors in an experimental evaluation of 19 forecasting models across several datasets. Finally, an additional ablation study determined that each component of the proposed solution enhances its overall performance.http://www.sciencedirect.com/science/article/pii/S2667096825000205Graph neural networksTraffic forecastingUrban mobility
spellingShingle Theodoros Theodoropoulos
Angelos-Christos Maroudis
Uwe Zdun
Antonios Makris
Konstantinos Tserpes
WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
International Journal of Information Management Data Insights
Graph neural networks
Traffic forecasting
Urban mobility
title WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
title_full WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
title_fullStr WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
title_full_unstemmed WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
title_short WEST GCN-LSTM: Weighted stacked spatio-temporal graph neural networks for regional traffic forecasting
title_sort west gcn lstm weighted stacked spatio temporal graph neural networks for regional traffic forecasting
topic Graph neural networks
Traffic forecasting
Urban mobility
url http://www.sciencedirect.com/science/article/pii/S2667096825000205
work_keys_str_mv AT theodorostheodoropoulos westgcnlstmweightedstackedspatiotemporalgraphneuralnetworksforregionaltrafficforecasting
AT angeloschristosmaroudis westgcnlstmweightedstackedspatiotemporalgraphneuralnetworksforregionaltrafficforecasting
AT uwezdun westgcnlstmweightedstackedspatiotemporalgraphneuralnetworksforregionaltrafficforecasting
AT antoniosmakris westgcnlstmweightedstackedspatiotemporalgraphneuralnetworksforregionaltrafficforecasting
AT konstantinostserpes westgcnlstmweightedstackedspatiotemporalgraphneuralnetworksforregionaltrafficforecasting