Analyzing the Role and Optimization of Loss Functions in Machine Learning and Deep Learning

Analyzing the Role and Optimization of Loss Functions in Machine Learning and Deep Learning | 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 Analyzing the Role and Optimization of Loss Functions in Machine Learning and Deep Learning Trinh Quang Minh, Ngo Thi Lan, Nguyen Minh Hieu, Tran Minh Tan, and 4 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-9673664/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 In the field of machine learning and deep learning, the loss function serves as a critical mathematical framework for quantifying the discrepancy between predicted outputs and ground-truth labels. This study provides a comprehensive analysis of loss function optimization across diverse and complex biological datasets, ranging from multi-label protein function prediction (CAFA 6) to high-dimensional RNA 3D structure modeling (Stanford RNA 3D Folding). Our methodology integrates advanced feature engineering—including K-mer tokenization, TF-IDF vectorization, and Truncated SVD—with specialized modeling strategies such as Template-Based Modeling (TBM) and sequence alignment. For categorical tasks, we demonstrate the efficacy of One-vs-Rest (OvR) strategies utilizing Binary Cross-Entropy (BCE) to manage large-scale Gene Ontology (GO) terms. In the domain of structural biology, the study explores the transition to spatial coordinate regression, where the optimization objective focuses on the precise positioning of C1' atoms to define molecular backbones. Experimental results, validated on Kaggle competition benchmarks, show that appropriate loss function selection and fine-tuning significantly enhance model convergence and prediction reliability, achieving a confidence threshold of over 0.85 for high-frequency biological annotations. This research underscores the pivotal role of loss functions in bridging the gap between raw genomic sequences and functional 3D molecular insights. Bioinformatics Loss Function Cost Function Machine Learning Deep Learning Optimization Ensemble Learning Kaggle Challenges Predictive Modeling 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-9673664","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":637847261,"identity":"d650390a-1381-4a9a-b34c-b83c3c948841","order_by":0,"name":"Trinh Quang Minh","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuElEQVRIiWNgGAWjYJCCA4wNNkASyOIhRjkPREuaBGlaGBgbDpOgxZ49O/Fw4Y7zdXw3EhgfvG1jSNxO0BaetxsOzzxzW0LyRgKz4Vyglp0NhLRI5G44zNt2W8LgRgKbNG8bg7HBAeK0nANpYf9NipYDYFuYgVrkCGs58xakJVly5pmHzZJzzkkQ1sLenrv5M2+bHT/f8eSDH96U2fAQ1MLAkABjMDYACQmC6pG1jIJRMApGwSjAAQBbdkIhjVUNywAAAABJRU5ErkJggg==","orcid":"https://orcid.org/0009-0000-5068-6774","institution":"Tay Do University","correspondingAuthor":true,"prefix":"","firstName":"Trinh","middleName":"Quang","lastName":"Minh","suffix":""},{"id":637847262,"identity":"f8810da8-5e5d-47c4-bdac-22a44e86791f","order_by":1,"name":"Ngo Thi Lan","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Ngo","middleName":"Thi","lastName":"Lan","suffix":""},{"id":637847263,"identity":"1fbcf915-125c-4dbc-9c5f-38e165e357b3","order_by":2,"name":"Nguyen Minh Hieu","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Nguyen","middleName":"Minh","lastName":"Hieu","suffix":""},{"id":637847264,"identity":"e7ae6989-534b-423b-a2d5-530b4f435cec","order_by":3,"name":"Tran Minh Tan","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Tran","middleName":"Minh","lastName":"Tan","suffix":""},{"id":637847265,"identity":"bad11cef-bd00-490d-ab10-42cfb1ca02ec","order_by":4,"name":"Do Chi Tam","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Do","middleName":"Chi","lastName":"Tam","suffix":""},{"id":637847266,"identity":"695af875-b2b1-4c67-a2b4-4e1cbae15ee6","order_by":5,"name":"Lam Tan Phuong6","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Lam","middleName":"Tan","lastName":"Phuong6","suffix":""},{"id":637847267,"identity":"fb2a7e7e-bbd9-4c64-87da-d188fe703904","order_by":6,"name":"Nguyen Chi Cuong","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Nguyen","middleName":"Chi","lastName":"Cuong","suffix":""},{"id":637847268,"identity":"83b29132-f4de-4875-905e-a41274bb47a0","order_by":7,"name":"Bui Xuan Tung","email":"","orcid":"","institution":"Tay Do University","correspondingAuthor":false,"prefix":"","firstName":"Bui","middleName":"Xuan","lastName":"Tung","suffix":""}],"badges":[],"createdAt":"2026-05-11 02:38:01","currentVersionCode":1,"declarations":{"humanSubjects":true,"vertebrateSubjects":false,"conflictsOfInterestStatement":false,"humanSubjectEthicalGuidelines":true,"humanSubjectConsent":true,"humanSubjectClinicalTrial":false,"humanSubjectCaseReport":false,"vertebrateSubjectEthicalGuidelines":false},"doi":"10.21203/rs.3.rs-9673664/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-9673664/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109102241,"identity":"1dd26406-f9ec-486c-bb3f-9d843a7f4f01","added_by":"auto","created_at":"2026-05-12 14:31:47","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":587951,"visible":true,"origin":"","legend":"","description":"","filename":"TrinhQuangMinh10h2712032026New01.pdf","url":"https://assets-eu.researchsquare.com/files/rs-9673664/v1_covered_e1d5ca01-7dbb-4125-90c1-3c7ac0d8807e.pdf"}],"financialInterests":"The authors declare no competing interests.","formattedTitle":"\u003cp\u003eAnalyzing the Role and Optimization of Loss Functions in Machine Learning and Deep Learning\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":"Loss Function, Cost Function, Machine Learning, Deep Learning, Optimization, Ensemble Learning, Kaggle Challenges, Predictive Modeling","lastPublishedDoi":"10.21203/rs.3.rs-9673664/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-9673664/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eIn the field of machine learning and deep learning, the loss function serves as a critical mathematical framework for quantifying the discrepancy between predicted outputs and ground-truth labels. This study provides a comprehensive analysis of loss function optimization across diverse and complex biological datasets, ranging from multi-label protein function prediction (CAFA 6) to high-dimensional RNA 3D structure modeling (Stanford RNA 3D Folding). Our methodology integrates advanced feature engineering\u0026mdash;including K-mer tokenization, TF-IDF vectorization, and Truncated SVD\u0026mdash;with specialized modeling strategies such as Template-Based Modeling (TBM) and sequence alignment. For categorical tasks, we demonstrate the efficacy of One-vs-Rest (OvR) strategies utilizing Binary Cross-Entropy (BCE) to manage large-scale Gene Ontology (GO) terms. In the domain of structural biology, the study explores the transition to spatial coordinate regression, where the optimization objective focuses on the precise positioning of C1' atoms to define molecular backbones. Experimental results, validated on Kaggle competition benchmarks, show that appropriate loss function selection and fine-tuning significantly enhance model convergence and prediction reliability, achieving a confidence threshold of over 0.85 for high-frequency biological annotations. This research underscores the pivotal role of loss functions in bridging the gap between raw genomic sequences and functional 3D molecular insights.\u003c/p\u003e","manuscriptTitle":"Analyzing the Role and Optimization of Loss Functions in Machine Learning and Deep Learning","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-12 14:30:34","doi":"10.21203/rs.3.rs-9673664/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":"24a0e46f-d120-4df9-a6ee-f50cc1143c1e","owner":[],"postedDate":"May 12th, 2026","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[{"id":68021429,"name":"Bioinformatics"}],"tags":[],"updatedAt":"2026-05-12T14:30:34+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-12 14:30:34","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-9673664","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-9673664","identity":"rs-9673664","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}