FreFilterTST: A Dynamic Channel Graph Sparsification Approach to Multivariate Time Series Anomaly Detection with Frequency-Domain Restoration

FreFilterTST: A Dynamic Channel Graph Sparsification Approach to Multivariate Time Series Anomaly Detection with Frequency-Domain Restoration | 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 Article FreFilterTST: A Dynamic Channel Graph Sparsification Approach to Multivariate Time Series Anomaly Detection with Frequency-Domain Restoration Yi Wang, Jian Jie Zhang, Ming Yang Zhang This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-7305484/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 23 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted 12 You are reading this latest preprint version Abstract The scope of time series anomaly detection is increasingly shifting from univariate to multivariate contexts, as a growing number of real-world problems can no longer be adequately addressed by analyzing individual variables in isolation. Consequently, multivariate time series anomaly detection is not only in high demand but also presents significant challenges. However, existing methods struggle with a dual challenge: capturing subtle, fine-grained frequency features, and effectively modeling complex inter-channel dependencies. Current channel-handling strategies are often either too restrictive, like Channel-Independent (CI) methods that ignore valuable correlations, or susceptible to noise, like Channel-Dependent (CD) methods that indiscriminately integrate all relationships.To address these challenges, we propose FreFilterTST, a novel framework that uniquely combines frequency-domain inpainting with adaptive Frequency-Amplitude filtering. Specifically, FreFilterTST first reconstructs the input sequence from spectral patches to establish a rich representation of normative patterns. Subsequently, a Transformer-based Mixture-of-Experts (MoE) architecture acts as an adaptive filter, dynamically identifying and preserving the most critical Frequency-Amplitude dependencies while pruning irrelevant or noisy ones. This can allow our model to overcome the inherent trade-offs of conventional CI and CD approaches.our Extensive experiments on six benchmark public datasets demonstrate that FreFilterTST achieves an excellent performance. Physical sciences/Engineering Physical sciences/Mathematics and computing Time Series Anomaly Detection Multivariate Time Series Frequency Patch Mixture of Experts (MoE) Full Text Additional Declarations No competing interests reported. Supplementary Files details.pdf Cite Share Download PDF Status: Published Journal Publication published 23 Dec, 2025 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 15 Oct, 2025 Reviews received at journal 14 Oct, 2025 Reviewers agreed at journal 14 Oct, 2025 Reviews received at journal 09 Oct, 2025 Reviewers agreed at journal 05 Oct, 2025 Reviews received at journal 02 Sep, 2025 Reviewers agreed at journal 19 Aug, 2025 Reviewers invited by journal 19 Aug, 2025 Editor assigned by journal 19 Aug, 2025 Editor invited by journal 12 Aug, 2025 Submission checks completed at journal 09 Aug, 2025 First submitted to journal 09 Aug, 2025 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-7305484","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":502623636,"identity":"be712f3f-6ff7-4685-a4a0-7dd2e8d49e1d","order_by":0,"name":"Yi Wang","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA00lEQVRIiWNgGAWjYBACxmYgIWHAIMfAcADIYiNBizHxWmAgsQFMEaOFuZ352QOLApv0DQfPGDB8KDvMwD+7gZDD2MwNJAzScmc2nDFgnHHuMIPEnQOEtDCYSUgYHM7tZzhjwMzbdpjBQCKBkBb2b0At/9PZQFr+EqeFB2TLgQR+kBZGIrWUAbUkG85sOFZwsOdcOo/EDQJaDPuPb5OW+GMnb3Dj8MYHP8qs5fhnENLSAAxoCRBL4gA4MnnwqwcCeZDjPoBY/A0EFY+CUTAKRsEIBQAKcz45RzwRCgAAAABJRU5ErkJggg==","orcid":"","institution":"Xinjiang University","correspondingAuthor":true,"prefix":"","firstName":"Yi","middleName":"","lastName":"Wang","suffix":""},{"id":502623638,"identity":"5eb6e02f-789f-4936-821d-ef6ed60ba68b","order_by":1,"name":"Jian Jie Zhang","email":"","orcid":"","institution":"Xinjiang University","correspondingAuthor":false,"prefix":"","firstName":"Jian","middleName":"Jie","lastName":"Zhang","suffix":""},{"id":502623639,"identity":"7fc094f9-504d-4243-b975-c2a12746804c","order_by":2,"name":"Ming Yang Zhang","email":"","orcid":"","institution":"Xinjiang University","correspondingAuthor":false,"prefix":"","firstName":"Ming","middleName":"Yang","lastName":"Zhang","suffix":""}],"badges":[],"createdAt":"2025-08-06 04:23:15","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-7305484/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-7305484/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-025-33186-1","type":"published","date":"2025-12-23T15:57:24+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":99172267,"identity":"f314adca-34db-40b0-8036-3830dcb28e92","added_by":"auto","created_at":"2025-12-29 16:06:44","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1206570,"visible":true,"origin":"","legend":"","description":"","filename":"FreFilterTST.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7305484/v1_covered_873ed9f6-3abc-4bc6-9795-7df6a42efe41.pdf"},{"id":89981772,"identity":"9edd2729-a9cb-4529-8802-801db2884ff7","added_by":"auto","created_at":"2025-08-27 06:26:27","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"supplement","size":50112,"visible":true,"origin":"","legend":"","description":"","filename":"details.pdf","url":"https://assets-eu.researchsquare.com/files/rs-7305484/v1/5ffedb991e17db38e6b25966.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"FreFilterTST: A Dynamic Channel Graph Sparsification Approach to Multivariate Time Series Anomaly Detection with Frequency-Domain Restoration","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":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Time Series Anomaly Detection, Multivariate Time Series, Frequency Patch, Mixture of Experts (MoE)","lastPublishedDoi":"10.21203/rs.3.rs-7305484/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-7305484/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"The scope of time series anomaly detection is increasingly shifting from univariate to multivariate contexts, as a growing number of real-world problems can no longer be adequately addressed by analyzing individual variables in isolation. Consequently, multivariate time series anomaly detection is not only in high demand but also presents significant challenges. However, existing methods struggle with a dual challenge: capturing subtle, fine-grained frequency features, and effectively modeling complex inter-channel dependencies. Current channel-handling strategies are often either too restrictive, like Channel-Independent (CI) methods that ignore valuable correlations, or susceptible to noise, like Channel-Dependent (CD) methods that indiscriminately integrate all relationships.To address these challenges, we propose FreFilterTST, a novel framework that uniquely combines frequency-domain inpainting with adaptive Frequency-Amplitude filtering. Specifically, FreFilterTST first reconstructs the input sequence from spectral patches to establish a rich representation of normative patterns. Subsequently, a Transformer-based Mixture-of-Experts (MoE) architecture acts as an adaptive filter, dynamically identifying and preserving the most critical Frequency-Amplitude dependencies while pruning irrelevant or noisy ones. This can allow our model to overcome the inherent trade-offs of conventional CI and CD approaches.our Extensive experiments on six benchmark public datasets demonstrate that FreFilterTST achieves an excellent performance.","manuscriptTitle":"FreFilterTST: A Dynamic Channel Graph Sparsification Approach to Multivariate Time Series Anomaly Detection with Frequency-Domain Restoration","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-08-27 06:26:23","doi":"10.21203/rs.3.rs-7305484/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-10-15T18:33:31+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-14T15:07:13+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"51536078747371145344400010416345584357","date":"2025-10-14T14:52:59+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-10-10T01:54:24+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"53025959344232397271399193515688720799","date":"2025-10-06T02:00:18+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-09-02T09:18:56+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"213573908241765711545696324052524820214","date":"2025-08-19T14:12:19+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-08-19T12:20:44+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-08-19T11:53:51+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2025-08-12T05:44:37+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-08-10T02:50:10+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2025-08-10T02:48:58+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"scientific-reports","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"scirep","sideBox":"Learn more about [Scientific Reports](http://www.nature.com/srep/)","snPcode":"","submissionUrl":"","title":"Scientific Reports","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"stoa","reportingPortfolio":"Scientific Reports","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"7892ce58-0aad-4af1-8175-bf68eeeb59ee","owner":[],"postedDate":"August 27th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":53392141,"name":"Physical sciences/Engineering"},{"id":53392142,"name":"Physical sciences/Mathematics and computing"}],"tags":[],"updatedAt":"2025-12-29T16:00:05+00:00","versionOfRecord":{"articleIdentity":"rs-7305484","link":"https://doi.org/10.1038/s41598-025-33186-1","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2025-12-23 15:57:24","publishedOnDateReadable":"December 23rd, 2025"},"versionCreatedAt":"2025-08-27 06:26:23","video":"","vorDoi":"10.1038/s41598-025-33186-1","vorDoiUrl":"https://doi.org/10.1038/s41598-025-33186-1","workflowStages":[]},"version":"v1","identity":"rs-7305484","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-7305484","identity":"rs-7305484","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}