Spatio-Temporal Crime Modeling in São Paulo: A Comparison of SAR and LSTM Approaches

Spatio-Temporal Crime Modeling in São Paulo: A Comparison of SAR and LSTM Approaches | 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 Spatio-Temporal Crime Modeling in São Paulo: A Comparison of SAR and LSTM Approaches Wellington Yuanhe Zhao, Cibele M. Russo, Luis Gustavo Nonato This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-8752288/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 This study investigates the prediction of urban crime in the São Paulo metropolitan region using an integrated spatio-temporal modeling framework. We compare Spatial Autoregressive (SAR) models with Artificial Neural Networks (ANN), including Long Short-Term Memory (LSTM) architectures, leveraging a comprehensive dataset of crime incidents, demographic indicators, and educational and economic variables at the municipal level. The SAR model captures spatial dependencies and allows for interpretable regional analysis, while LSTM networks model complex nonlinear patterns and temporal dynamics. Empirical results show that although the SAR model explains crime variation effectively, LSTM achieves substantially better predictive performance (RMSE: 22.64 vs 77.18 forSAR). This suggests the potential of deep learning techniques in improving urban crime forecasting. Our findings highlight the importance of combining spatial econometrics and machine learning to support evidence-based public safety policies and targeted social interventions. crime modeling spatial data temporal data Full Text Additional Declarations No competing interests reported. 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-8752288","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":637328371,"identity":"c4e4e1b1-0e60-4bdc-92e5-cdb8c7db4b59","order_by":0,"name":"Wellington Yuanhe Zhao","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABFklEQVRIiWNgGAWjYFAC5sYDCQjeAQZ+EJVQgE8LYwNIiwRci2QDSIsBAS0MyFoMQFwGPFoMjjc2HHi4o7aOX7rHTOLnjjtyxudXJ354YMAgzy92ALuWMwcbDiSeOS4hOeeMmWTvmWfGZjfebpYAOsxw5uwErFrMbiQCtbQdkzC4kWMmwdt2OHHbjbMbQFoSDG7j0HL/IUKL5N+2w/WbZ5zd/AOvlhuMIC01YC3SQFsSDPh7t+G1xf4M2GEHJGfOSCu2lm07bDjjBu82iwQDCZx+kWw/fPDhz7Y6fn6J5I0337YdlufvP7v55o8KG3l+aexaoOAwEHNA40ICrFICt2IIqANi9gcQNv8BQqpHwSgYBaNghAEAAnRufWNH9h4AAAAASUVORK5CYII=","orcid":"","institution":"Universidade de São Paulo","correspondingAuthor":true,"prefix":"","firstName":"Wellington","middleName":"Yuanhe","lastName":"Zhao","suffix":""},{"id":637328372,"identity":"f4de1701-4fa6-41a9-bb08-28492bfa0d87","order_by":1,"name":"Cibele M. Russo","email":"","orcid":"","institution":"Universidade de São Paulo","correspondingAuthor":false,"prefix":"","firstName":"Cibele","middleName":"M.","lastName":"Russo","suffix":""},{"id":637328373,"identity":"27465e1b-605e-4705-950d-76ccb909e22e","order_by":2,"name":"Luis Gustavo Nonato","email":"","orcid":"","institution":"Universidade de São Paulo","correspondingAuthor":false,"prefix":"","firstName":"Luis","middleName":"Gustavo","lastName":"Nonato","suffix":""}],"badges":[],"createdAt":"2026-01-31 19:53:32","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-8752288/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-8752288/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":109027991,"identity":"273fda98-83ea-4f0e-b10f-0c41e216a34b","added_by":"auto","created_at":"2026-05-11 22:11:14","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1264531,"visible":true,"origin":"","legend":"","description":"","filename":"wellingtonpaper1.pdf","url":"https://assets-eu.researchsquare.com/files/rs-8752288/v1_covered_27d063dd-6cda-4999-adf4-d16cd49639ed.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"\u003cp\u003eSpatio-Temporal Crime Modeling in São Paulo: A Comparison of SAR and LSTM Approaches\u003c/p\u003e","fulltext":[],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":false,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"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":"crime modeling, spatial data, temporal data","lastPublishedDoi":"10.21203/rs.3.rs-8752288/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-8752288/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"This study investigates the prediction of urban crime in the São Paulo metropolitan region using an integrated spatio-temporal modeling framework. We compare Spatial Autoregressive (SAR) models with Artificial Neural Networks (ANN), including Long Short-Term Memory (LSTM) architectures, leveraging a comprehensive dataset of crime incidents, demographic indicators, and educational and economic variables at the municipal level. The SAR model captures spatial dependencies and allows for interpretable regional analysis, while LSTM networks model complex nonlinear patterns and temporal dynamics. Empirical results show that although the SAR model explains crime variation effectively, LSTM achieves substantially better predictive performance (RMSE: 22.64 vs 77.18 forSAR). This suggests the potential of deep learning techniques in improving urban crime forecasting. Our findings highlight the importance of combining spatial econometrics and machine learning to support evidence-based public safety policies and targeted social interventions.","manuscriptTitle":"Spatio-Temporal Crime Modeling in São Paulo: A Comparison of SAR and LSTM Approaches","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2026-05-11 22:11:00","doi":"10.21203/rs.3.rs-8752288/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":"919eb554-b211-4f65-8054-796472a36148","owner":[],"postedDate":"May 11th, 2026","published":true,"recentEditorialEvents":[{"type":"editorInvitedReview","content":"","date":"2026-05-09T10:33:52+00:00","index":13,"fulltext":""},{"type":"reviewerAgreed","content":"325782427148146094637381831056897966914","date":"2026-05-04T17:03:30+00:00","index":11,"fulltext":""},{"type":"reviewersInvited","content":"7","date":"2026-05-04T00:30:47+00:00","index":"","fulltext":""}],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-05-11T22:11:00+00:00","versionOfRecord":[],"versionCreatedAt":"2026-05-11 22:11:00","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-8752288","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-8752288","identity":"rs-8752288","version":["v1"]},"buildId":"XKTyCvWXoU3ODBz1xrDgd","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}