Optimal Adaptive Curriculum Generation for Continual Learning via Multi-Objective Variational Optimization

Optimal Adaptive Curriculum Generation for Continual Learning via Multi-Objective Variational Optimization | 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 Optimal Adaptive Curriculum Generation for Continual Learning via Multi-Objective Variational Optimization Weihao Li, Jingxuan Sun, Meilin Qiu, Zhenyu Gao, Zhou Shen This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8363440/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The increasing demand for intelligent systems capable of continual learning necessitates effective strategies for adaptive curriculum generation. In this paper, we introduce a novel optimization framework that addresses the dual objectives of maximizing forward transfer and minimizing catastrophic forgetting within resource constraints. We formulate the adaptive curriculum generation problem as a multi-objective variational optimization challenge, presenting an innovative algorithm that leverages gradient-based methods for efficient solution discovery. Our contributions include a comprehensive theoretical analysis that provides guarantees on generalization, convergence, and efficiency, alongside extensive empirical evaluations across standard benchmarks such as Split MNIST and CIFAR-100. Results demonstrate that our approach outperforms state-of-the-art methods, exhibiting superior forward transfer rates and reduced forgetting while optimizing resource utilization. By establishing new benchmarks for curriculum quality in the context of continual learning, we pave the way for future research and application of adaptive learning strategies across diverse domains. The findings underscore the importance of strategic curriculum design in enhancing learning outcomes and advancing autonomous AI systems. continual learning curriculum optimization multi-objective optimization variational inference catastrophic forgetting generalization resource constraints Full Text Additional Declarations The authors declare no competing interests. Cite Share Download PDF Status: Posted Version 1 posted 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-8363440","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":560377978,"identity":"0c07042e-c440-411e-8d77-db622a90dd1b","order_by":0,"name":"Weihao Li","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Weihao","middleName":"","lastName":"Li","suffix":""},{"id":560377979,"identity":"3c71f801-2956-4829-91e5-34604135128c","order_by":1,"name":"Jingxuan Sun","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Jingxuan","middleName":"","lastName":"Sun","suffix":""},{"id":560377980,"identity":"93ed64ad-f9c3-4579-ade0-86fbe26071c5","order_by":2,"name":"Meilin Qiu","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Meilin","middleName":"","lastName":"Qiu","suffix":""},{"id":560377981,"identity":"2d891cfa-f291-4ae9-9192-8344ce0cf618","order_by":3,"name":"Zhenyu Gao","email":"","orcid":"","institution":"","correspondingAuthor":false,"prefix":"","firstName":"Zhenyu","middleName":"","lastName":"Gao","suffix":""},{"id":560377982,"identity":"066ace46-af95-4e1b-8861-ecaa344ed44d","order_by":4,"name":"Zhou Shen","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFElEQVRIiWNgGAWjYDACCSBmbEASMGAH8QwsSNHCcwBESpCiRSIBJo4dyM9uPvbw5w6bxAbpw4c/fKi5J28u+fzqhh8FEgz87d0J2LQY3DmWbiB5Ji2xgS8tTXLGsWLDnbNzym72AB0mcebsBqxaJHLMJAzbDic28PCYMfOwJSQY3M5Ju8ED1GIgkYtVi/yM/G8SiWAt/J8///kH1HLzTNrNP3i0MNzIYZM4CLGFQZqxDajlBvux2/hsMbiRZibZ2JZm3MbDZibZ25dguOFMDtttGQMJHlx+kZ+R/EzyZ5uNbD8P8+MPP74lyBscP/7s5ps/NnL87b3YHQYDbAgmjwGYxKscDbA/IEX1KBgFo2AUDH8AAA8DYpsNUdNBAAAAAElFTkSuQmCC","orcid":"","institution":"","correspondingAuthor":true,"prefix":"","firstName":"Zhou","middleName":"","lastName":"Shen","suffix":""}],"badges":[],"createdAt":"2025-12-15 08:09:02","currentVersionCode":1,"declarations":{"humanSubjects":false,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":false,"humanSubjectConsent":false,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-8363440/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8363440/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":98623813,"identity":"84fc007b-367d-4ad5-8556-18b6c3e00d00","added_by":"auto","created_at":"2025-12-19 17:07:38","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":231214,"visible":true,"origin":"","legend":"","description":"","filename":"paper.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8363440/v1_covered_b8b56313-b9e1-49c2-9904-9cd52f231b55.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eOptimal Adaptive Curriculum Generation for Continual Learning via Multi-Objective Variational Optimization\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":true,"hideJournal":true,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":true,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":true,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"continual learning, curriculum optimization, multi-objective optimization, variational inference, catastrophic forgetting, generalization, resource constraints","lastPublishedDoi":"10.21203/rs.3.rs-8363440/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8363440/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe increasing demand for intelligent systems capable of continual learning necessitates effective strategies for adaptive curriculum generation. In this paper, we introduce a novel optimization framework that addresses the dual objectives of maximizing forward transfer and minimizing catastrophic forgetting within resource constraints. We formulate the adaptive curriculum generation problem as a multi-objective variational optimization challenge, presenting an innovative algorithm that leverages gradient-based methods for efficient solution discovery. Our contributions include a comprehensive theoretical analysis that provides guarantees on generalization, convergence, and efficiency, alongside extensive empirical evaluations across standard benchmarks such as Split MNIST and CIFAR-100. Results demonstrate that our approach outperforms state-of-the-art methods, exhibiting superior forward transfer rates and reduced forgetting while optimizing resource utilization. By establishing new benchmarks for curriculum quality in the context of continual learning, we pave the way for future research and application of adaptive learning strategies across diverse domains. The findings underscore the importance of strategic curriculum design in enhancing learning outcomes and advancing autonomous AI systems.\u003c/p\u003e","manuscriptTitle":"Optimal Adaptive Curriculum Generation for Continual Learning via Multi-Objective Variational Optimization","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-12-18 04:37:49","doi":"10.21203/rs.3.rs-8363440/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"ed81d30f-bcb1-4766-8ad1-01209863f3ff","owner":[],"postedDate":"December 18th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2025-12-18T04:37:49+00:00","versionOfRecord":[],"versionCreatedAt":"2025-12-18 04:37:49","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8363440","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8363440","identity":"rs-8363440","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}