A Hybrid Ontology-Based Feature SelectionFramework for Enhancing Predictive Accuracy inRegression Models
preprint
OA: closed
AI-generated deep summary
This paper studied how to improve regression-model predictive accuracy for “firefighter interventions” using hybrid feature selection that combines ontology-based reasoning with machine learning. The authors developed a domain-specific ontology capturing environmental, temporal, and intervention-related factors, then used ontology-derived centrality metrics (degree, closeness, betweenness) alongside ML feature selection for models trained with XGBoost, LightGBM, and LSTM, comparing results against ML-only feature selection. Across models, the hybrid approach consistently improved performance (e.g., XGBoost R² 0.976 vs 0.97; LightGBM R² 0.975 vs 0.97; LSTM R² 0.964 vs 0.96), and the authors emphasize increased interpretability and context relevance. The preprint explicitly notes it is not peer reviewed. The paper does not explicitly discuss endometriosis or adenomyosis; it was included in the corpus via a keyword match in the upstream search index.
Read from the paper's body, not the abstract. Not a substitute for reading the paper. No clinical advice. How this works
Abstract
Abstract Predicting firefighter interventions presents a complex challenge due to the high dimensionality and intricacy of the data. While machine learning (ML) technologies offer promising solutions, ineffective feature selection can significantly hinder model performance and reduce predictive accuracy. This study proposes a hybrid feature selection approach that combines ontology-based reasoning with traditional ML techniques to enhance the predictive accuracy of regression models for firefighter interventions. We utilized three machine learning algorithms—XGBoost, LightGBM, and Long Short-Term Memory (LSTM) networks—across two feature selection strategies: one solely based on ML algorithms, and another using a hybrid approach that integrates ontology-based centrality metrics, such as degree, closeness, and betweenness, with ML techniques. A domain-specific ontology was developed to capture key environmental, temporal, and intervention-related factors, improving the feature selection process for more interpretable and contextually relevant features. The results clearly show that the hybrid feature selection approach consistently outperforms the ML-only method. For the XGBoost model, the hybrid approach resulted in an R2 of 0.976, compared to 0.97 for the ML-only method. The LSTM model also saw improvements, with the hybrid approach achieving an R2 of 0.964, compared to 0.96 for ML-only. Similarly, for the LightGBM model, the hybrid approach produced an R2 of 0.975, compared to 0.97 for ML-only. This research underscores the significant advantages of combining ontology-based feature selection with ML, leading to improved predictive accuracy and better model interpretability, particularly in high-dimensional data environments.
Full text
15,087 characters
· extracted from
preprint-html
· click to expand
A Hybrid Ontology-Based Feature SelectionFramework for Enhancing Predictive Accuracy inRegression Models | 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 A Hybrid Ontology-Based Feature SelectionFramework for Enhancing Predictive Accuracy inRegression Models Sarah Ayad, Roxane Mallouhy, Christophe Guyeux This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-5325338/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 18 Jun, 2025 Read the published version in Knowledge and Information Systems → Version 1 posted 15 You are reading this latest preprint version Abstract Predicting firefighter interventions presents a complex challenge due to the high dimensionality and intricacy of the data. While machine learning (ML) technologies offer promising solutions, ineffective feature selection can significantly hinder model performance and reduce predictive accuracy. This study proposes a hybrid feature selection approach that combines ontology-based reasoning with traditional ML techniques to enhance the predictive accuracy of regression models for firefighter interventions. We utilized three machine learning algorithms—XGBoost, LightGBM, and Long Short-Term Memory (LSTM) networks—across two feature selection strategies: one solely based on ML algorithms, and another using a hybrid approach that integrates ontology-based centrality metrics, such as degree, closeness, and betweenness, with ML techniques. A domain-specific ontology was developed to capture key environmental, temporal, and intervention-related factors, improving the feature selection process for more interpretable and contextually relevant features. The results clearly show that the hybrid feature selection approach consistently outperforms the ML-only method. For the XGBoost model, the hybrid approach resulted in an R 2 of 0.976, compared to 0.97 for the ML-only method. The LSTM model also saw improvements, with the hybrid approach achieving an R 2 of 0.964, compared to 0.96 for ML-only. Similarly, for the LightGBM model, the hybrid approach produced an R 2 of 0.975, compared to 0.97 for ML-only. This research underscores the significant advantages of combining ontology-based feature selection with ML, leading to improved predictive accuracy and better model interpretability, particularly in high-dimensional data environments. Hybrid Methodology Domain Ontology Centrality Metrics Feature selection Firefighter Intervention Prediction XGBoost LightGBM LSTM Predictive Accuracy Ontology-Based Feature Selection Full Text Additional Declarations No competing interests reported. Cite Share Download PDF Status: Published Journal Publication published 18 Jun, 2025 Read the published version in Knowledge and Information Systems → Version 1 posted Editorial decision: Revision requested 16 Dec, 2024 Reviews received at journal 04 Dec, 2024 Reviewers agreed at journal 29 Nov, 2024 Reviews received at journal 26 Nov, 2024 Reviews received at journal 18 Nov, 2024 Reviews received at journal 15 Nov, 2024 Reviewers agreed at journal 15 Nov, 2024 Reviewers agreed at journal 15 Nov, 2024 Reviewers agreed at journal 12 Nov, 2024 Reviewers agreed at journal 12 Nov, 2024 Reviewers agreed at journal 10 Nov, 2024 Reviewers invited by journal 10 Nov, 2024 Editor assigned by journal 07 Nov, 2024 Submission checks completed at journal 26 Oct, 2024 First submitted to journal 24 Oct, 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-5325338","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":372017061,"identity":"083765e7-f186-468c-83bf-90244b361494","order_by":0,"name":"Sarah Ayad","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA4klEQVRIiWNgGAWjYNACNhsgwQNmMjYQqSWNdC2HSdAi38D8TOJH2Xl58/azBx/zMNjIbjjA/vADPi0GB9jMJHvO3TaccyYv2ZiHIc14wwEeYwm8WhgYzCR4224zzmDIMZPmYTicCNTCgFeLfAP7N8m/befsZ/C/AWn5D9TC/vgHXs8c4DGT5m07kDhDAmzLAaAWoL14HXaYp9ha5lxy8gyJd8mGcwySjWce5jGzwOuw9vaNN9+U2dnO4M89+OBNhZ1s3/H2xzfwOoyZgQXJGQZgEYKAGW8sjIJRMApGwShgAAB2aEU7PQFBBAAAAABJRU5ErkJggg==","orcid":"","institution":"Arab Open University","correspondingAuthor":true,"prefix":"","firstName":"Sarah","middleName":"","lastName":"Ayad","suffix":""},{"id":372017062,"identity":"a6f00fb8-444d-4b57-abe1-fb682c63e415","order_by":1,"name":"Roxane Mallouhy","email":"","orcid":"","institution":"Al Yamamah University","correspondingAuthor":false,"prefix":"","firstName":"Roxane","middleName":"","lastName":"Mallouhy","suffix":""},{"id":372017063,"identity":"a54e5aec-0ccb-4cdf-9a76-fb58e1719a6c","order_by":2,"name":"Christophe Guyeux","email":"","orcid":"","institution":"University of Franche-Comté","correspondingAuthor":false,"prefix":"","firstName":"Christophe","middleName":"","lastName":"Guyeux","suffix":""}],"badges":[],"createdAt":"2024-10-24 11:08:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-5325338/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-5325338/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1007/s10115-025-02497-0","type":"published","date":"2025-06-18T15:57:53+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":85231431,"identity":"d38ca5cf-97a3-40ba-9ed3-f6635e7ccc17","added_by":"auto","created_at":"2025-06-23 16:07:53","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1524182,"visible":true,"origin":"","legend":"","description":"","filename":"EnhancingPredictiveAccuracyinRegressionModelsthroughOntologyBasedFeatureSelection2.pdf","url":"https://assets-eu.researchsquare.com/files/rs-5325338/v1_covered_3ff28c78-41ca-4173-9c60-78eb88d54351.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"A Hybrid Ontology-Based Feature SelectionFramework for Enhancing Predictive Accuracy inRegression Models","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":"knowledge-and-information-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"kais","sideBox":"Learn more about [Knowledge and Information Systems](http://link.springer.com/journal/10115)","snPcode":"10115","submissionUrl":"https://submission.nature.com/new-submission/10115/3","title":"Knowledge and Information Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false},"keywords":"Hybrid Methodology, Domain Ontology, Centrality Metrics, Feature selection, Firefighter Intervention Prediction, XGBoost, LightGBM, LSTM, Predictive Accuracy, Ontology-Based Feature Selection","lastPublishedDoi":"10.21203/rs.3.rs-5325338/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-5325338/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e Predicting firefighter interventions presents a complex challenge due to the high dimensionality and intricacy of the data. While machine learning (ML) technologies offer promising solutions, ineffective feature selection can significantly hinder model performance and reduce predictive accuracy. This study proposes a hybrid feature selection approach that combines ontology-based reasoning with traditional ML techniques to enhance the predictive accuracy of regression models for firefighter interventions. We utilized three machine learning algorithms\u0026mdash;XGBoost, LightGBM, and Long Short-Term Memory (LSTM) networks\u0026mdash;across two feature selection strategies: one solely based on ML algorithms, and another using a hybrid approach that integrates ontology-based centrality metrics, such as degree, closeness, and betweenness, with ML techniques. A domain-specific ontology was developed to capture key environmental, temporal, and intervention-related factors, improving the feature selection process for more interpretable and contextually relevant features. The results clearly show that the hybrid feature selection approach consistently outperforms the ML-only method. For the XGBoost model, the hybrid approach resulted in an R\u003csup\u003e2\u003c/sup\u003e of 0.976, compared to 0.97 for the ML-only method. The LSTM model also saw improvements, with the hybrid approach achieving an R\u003csup\u003e2\u003c/sup\u003e of 0.964, compared to 0.96 for ML-only. Similarly, for the LightGBM model, the hybrid approach produced an R\u003csup\u003e2\u003c/sup\u003e of 0.975, compared to 0.97 for ML-only. This research underscores the significant advantages of combining ontology-based feature selection with ML, leading to improved predictive accuracy and better model interpretability, particularly in high-dimensional data environments.\u003c/p\u003e","manuscriptTitle":"A Hybrid Ontology-Based Feature SelectionFramework for Enhancing Predictive Accuracy inRegression Models","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-11-06 08:42:33","doi":"10.21203/rs.3.rs-5325338/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-12-17T02:17:47+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-12-04T06:21:25+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"125215746481644072571942773459921686761","date":"2024-11-29T07:39:35+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-26T19:10:10+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-18T07:09:42+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-11-15T21:40:29+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"244070904818777660337074685468690519132","date":"2024-11-15T11:27:59+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"234532166847321217231649206655689430251","date":"2024-11-15T06:50:30+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"231476777358113590235820436365799458427","date":"2024-11-12T09:51:49+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"36925964666706541855293313524270584325","date":"2024-11-12T06:24:19+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"173567801366615172161117792549837280354","date":"2024-11-10T18:36:02+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-11-10T06:05:45+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-11-07T07:35:03+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-10-26T07:17:47+00:00","index":"","fulltext":""},{"type":"submitted","content":"Knowledge and Information Systems","date":"2024-10-24T11:03:56+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"knowledge-and-information-systems","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":false,"externalIdentity":"kais","sideBox":"Learn more about [Knowledge and Information Systems](http://link.springer.com/journal/10115)","snPcode":"10115","submissionUrl":"https://submission.nature.com/new-submission/10115/3","title":"Knowledge and Information Systems","twitterHandle":"","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"em","reportingPortfolio":"Springer Hybrid","inReviewEnabled":true,"inReviewRevisionsEnabled":false}}],"origin":"","ownerIdentity":"7c09b4c2-ce3a-4961-adac-ac71d6ee5fb0","owner":[],"postedDate":"November 6th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[],"tags":[],"updatedAt":"2025-06-23T16:02:47+00:00","versionOfRecord":{"articleIdentity":"rs-5325338","link":"https://doi.org/10.1007/s10115-025-02497-0","journal":{"identity":"knowledge-and-information-systems","isVorOnly":false,"title":"Knowledge and Information Systems"},"publishedOn":"2025-06-18 15:57:53","publishedOnDateReadable":"June 18th, 2025"},"versionCreatedAt":"2024-11-06 08:42:33","video":"","vorDoi":"10.1007/s10115-025-02497-0","vorDoiUrl":"https://doi.org/10.1007/s10115-025-02497-0","workflowStages":[]},"version":"v1","identity":"rs-5325338","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-5325338","identity":"rs-5325338","version":["v1"]},"buildId":"qtupq5eGEP_6zYnWcrvyt","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}
Text is read by the "Ask this paper" AI Q&A widget below. Extraction quality varies by source — PMC NXML preserves structure cleanly, OA-HTML may include some navigation residue, and OA-PDF can have broken hyphenation. The publisher copy (via DOI) is the canonical version.
My notes (saved in your browser only)
Ask this paper
Answers must be backed by verbatim quotes from this paper's full text. Hallucinated quotes are dropped automatically; if no verbatim passage answers the question, we say so. How this works
Citation neighborhood (no data yet)
We don't have any in-corpus citations linked to this paper yet. This is a recent paper (2024) — citers typically take a year or two to land, and the OpenAlex reference graph may still be filling in.
Source provenance
- europepmc
- last seen: 2026-05-20T01:45:00.602351+00:00