Stacking multi-timescale graph attention network and bidirectional gated recurrent units for traffic flow prediction

Stacking multi-timescale graph attention network and bidirectional gated recurrent units for traffic flow prediction | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Research Article Stacking multi-timescale graph attention network and bidirectional gated recurrent units for traffic flow prediction Yizhi Liu, Zhengbiao Zou, Jingxin Tang, Zhuhua Liao, Yuxuan Liu, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5381241/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 08 Sep, 2025 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted 15 You are reading this latest preprint version Abstract Accurately predicting traffic flow becomes more and more important to intelligent transportation systems in the age of autonomous driving. Some existing methods extract spatial and temporal features of road networks respectively via graph convolutional network and recurrent neural network for traffic flow prediction. However, there are still threefold challenge of spatiotemporal coupling interrelations, dynamic spatial correlations, and the impact of periodic patterns and external factors. Addressing at these challenges, this paper proposes a novel approach of stacking multi-timescale graph attention network (GAT) and bidirectional gated recurrent units (BiGRU) for traffic flow prediction, called SMGAT-BiGRU. Firstly, a GAT feature extraction module is explored to capture the dynamic spatial correlation of different road segments, and then a multi-timescale feature fusion module is constructed to mine hidden periodic patterns and to capture the impact of external factors by extracting fine-grained spa-tiotemporal features and external dynamic time-varying features. Secondly, a BiGRU module with attention is proposed to capture temporal dependencies from dynamic spatial dependency sequences. Furthermore, a GAT-BiGRU module is presented to be stacked to fully capture the spatiotemporal coupling interrelations. Experimental results on two PeMS datasets show that SMGAT-BiGRU 1 can remarkably improve the performance of traffic flow prediction and outper-forms many state-of-the-art methods (such as T-GCN, STGCN, STSGCN, and DMSTGCN). Intelligent transportation system Traffic flow predictio Spatiotemporal data mining Graph attention network (GAT) Gated recurrent units (GRU) Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 08 Sep, 2025 Read the published version in International Journal of Machine Learning and Cybernetics → Version 1 posted Editorial decision: Revision requested 17 Dec, 2024 Reviews received at journal 10 Dec, 2024 Reviews received at journal 09 Dec, 2024 Reviewers agreed at journal 09 Dec, 2024 Reviews received at journal 03 Dec, 2024 Reviews received at journal 02 Dec, 2024 Reviewers agreed at journal 12 Nov, 2024 Reviews received at journal 11 Nov, 2024 Reviewers agreed at journal 08 Nov, 2024 Reviewers agreed at journal 07 Nov, 2024 Reviewers agreed at journal 06 Nov, 2024 Reviewers invited by journal 06 Nov, 2024 Editor assigned by journal 05 Nov, 2024 Submission checks completed at journal 04 Nov, 2024 First submitted to journal 03 Nov, 2024 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-5381241","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":377034577,"identity":"7fb56b80-c202-49ed-97ba-9b5e06d14ee8","order_by":0,"name":"Yizhi Liu","email":"","orcid":"","institution":"Hunan University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yizhi","middleName":"","lastName":"Liu","suffix":""},{"id":377034578,"identity":"90011405-ebd1-4d4d-b957-83c57fdc493d","order_by":1,"name":"Zhengbiao Zou","email":"","orcid":"","institution":"Hunan University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Zhengbiao","middleName":"","lastName":"Zou","suffix":""},{"id":377034579,"identity":"070a5d80-927c-4b31-9172-a996cb17d1a7","order_by":2,"name":"Jingxin Tang","email":"","orcid":"","institution":"Hunan University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Jingxin","middleName":"","lastName":"Tang","suffix":""},{"id":377034580,"identity":"0b03cd1c-958f-4ed5-8d91-969d2860ccff","order_by":3,"name":"Zhuhua Liao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYDACCeaGAwwMNglQLjMxWhhBWtJAWhgbiNYCJA+ToEV+dmPjgR8V5/PMJdKfP2CosE5sYD97AK8WxjkHGw72nLldbDkjx7CB4Ux6YgNPXgJeLcwSiQ0HeNtuJ264kcPYwNh2OLFBgscArxY2oJaDf9vOAbWkP2xg/EeEFh6glsO8bQeAWhIMGxgbiNAiAdIicyY5ccOZN4YzEo6lG7fx5ODXIj8j+fDHNxV2iRuOpz/48KHGWraf/Qx+LaggAeQ7EtSPglEwCkbBKMABAMjUTW0DGX97AAAAAElFTkSuQmCC","orcid":"","institution":"Hunan University of Science and Technology","correspondingAuthor":true,"prefix":"","firstName":"Zhuhua","middleName":"","lastName":"Liao","suffix":""},{"id":377034581,"identity":"1803187c-4737-4925-87fd-18b01c6cf247","order_by":4,"name":"Yuxuan Liu","email":"","orcid":"","institution":"Hunan University of Science and Technology","correspondingAuthor":false,"prefix":"","firstName":"Yuxuan","middleName":"","lastName":"Liu","suffix":""},{"id":377034582,"identity":"ccdbd031-e354-42ad-ab5a-751d45b91a43","order_by":5,"name":"Zhixiong Fang","email":"","orcid":"","institution":"Central Hospital of Xiangtan","correspondingAuthor":false,"prefix":"","firstName":"Zhixiong","middleName":"","lastName":"Fang","suffix":""}],"badges":[],"createdAt":"2024-11-03 08:53:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5381241/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5381241/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s13042-025-02783-x","type":"published","date":"2025-09-08T15:57:12+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":91359060,"identity":"db00838c-9719-4814-8b36-2c1dca318859","added_by":"auto","created_at":"2025-09-15 16:04:42","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1159943,"visible":true,"origin":"","legend":"","description":"","filename":"snarticle.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5381241/v1_covered_4d4222bf-2ac8-4e62-a02d-ee64b78e1583.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Stacking multi-timescale graph attention network and bidirectional gated recurrent units for traffic flow prediction","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":true,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Intelligent transportation system, Traffic flow predictio, Spatiotemporal data mining, Graph attention network (GAT), Gated recurrent units (GRU)","lastPublishedDoi":"10.21203/rs.3.rs-5381241/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5381241/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"Accurately predicting traffic flow becomes more and more important to intelligent transportation systems in the age of autonomous driving. Some existing methods extract spatial and temporal features of road networks respectively via graph convolutional network and recurrent neural network for traffic flow prediction. However, there are still threefold challenge of spatiotemporal coupling interrelations, dynamic spatial correlations, and the impact of periodic patterns and external factors. Addressing at these challenges, this paper proposes a novel approach of stacking multi-timescale graph attention network (GAT) and bidirectional gated recurrent units (BiGRU) for traffic flow prediction, called SMGAT-BiGRU. Firstly, a GAT feature extraction module is explored to capture the dynamic spatial correlation of different road segments, and then a multi-timescale feature fusion module is constructed to mine hidden periodic patterns and to capture the impact of external factors by extracting fine-grained spa-tiotemporal features and external dynamic time-varying features. Secondly, a BiGRU module with attention is proposed to capture temporal dependencies from dynamic spatial dependency sequences. Furthermore, a GAT-BiGRU module is presented to be stacked to fully capture the spatiotemporal coupling interrelations. Experimental results on two PeMS datasets show that SMGAT-BiGRU 1 can remarkably improve the performance of traffic flow prediction and outper-forms many state-of-the-art methods (such as T-GCN, STGCN, STSGCN, and DMSTGCN).","manuscriptTitle":"Stacking multi-timescale graph attention network and bidirectional gated recurrent units for traffic flow prediction","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-18 11:34:39","doi":"10.21203/rs.3.rs-5381241/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-12-17T07:31:28+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-10T13:33:23+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-09T08:18:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"238566916854402002185013898045803300067","date":"2024-12-09T05:07:08+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-03T06:44:15+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-02T12:58:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"188835491083780487216306278344187816890","date":"2024-11-12T08:06:20+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-11T14:34:15+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"331555050037013964981769599482138518658","date":"2024-11-09T02:17:35+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"65886997992375000291530865543415219343","date":"2024-11-07T06:13:03+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"126475116941665456421770525474146146360","date":"2024-11-07T02:58:49+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-07T01:20:52+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-06T01:01:57+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-11-05T03:02:15+00:00","index":"","fulltext":""},{"type":"submitted","content":"International Journal of Machine Learning and Cybernetics","date":"2024-11-03T08:48:53+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"international-journal-of-machine-learning-and-cybernetics","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"jmlc","sideBox":"Learn more about [International Journal of Machine Learning and Cybernetics](http://actavetscand.biomedcentral.com/)","snPcode":"13042","submissionUrl":"https://submission.nature.com/new-submission/13042/3","title":"International Journal of Machine Learning and Cybernetics","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"de968f86-f52b-458f-b836-4ef87ddd4be7","owner":[],"postedDate":"November 18th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-09-15T16:01:28+00:00","versionOfRecord":{"articleIdentity":"rs-5381241","link":"https://doi.org/10.1007/s13042-025-02783-x","journal":{"identity":"international-journal-of-machine-learning-and-cybernetics","isVorOnly":false,"title":"International Journal of Machine Learning and Cybernetics"},"publishedOn":"2025-09-08 15:57:12","publishedOnDateReadable":"September 8th, 2025"},"versionCreatedAt":"2024-11-18 11:34:39","video":"","vorDoi":"10.1007/s13042-025-02783-x","vorDoiUrl":"https://doi.org/10.1007/s13042-025-02783-x","workflowStages":[]},"version":"v1","identity":"rs-5381241","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5381241","identity":"rs-5381241","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}