Beyond Human Bias: Learning from CNNs’ Superior Insights in Satellite-Based Poverty Mapping

Beyond Human Bias: Learning from CNNs’ Superior Insights in Satellite-Based Poverty Mapping | 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 Beyond Human Bias: Learning from CNNs’ Superior Insights in Satellite-Based Poverty Mapping Hamid Sarmadi, Ibrahim Wahab, Ola Hall, Thorsteinn Rögnvaldsson, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-4545272/v1 This work is licensed under a CC BY 4.0 License Status: Published Journal Publication published 02 Oct, 2024 Read the published version in Scientific Reports → Version 1 posted 10 You are reading this latest preprint version Abstract Background : Satellite imagery is a potent tool for estimating human wealth and poverty, especially in regions lacking reliable data. This study compares a range of poverty estimation approaches from satellite images, spanning from expert-based to fully machine learning-based methodologies. Methods : Human experts ranked clusters from the DHS survey using high-resolution satellite images. Then expert-defined features were utilized in a machine learning algorithm to estimate poverty. An explainability method was applied to assess the importance and interaction of these features in poverty prediction. Additionally, a convolutional neural network (CNN) was employed to estimate poverty from low-resolution satellite images of the same locations. Findings : Our analysis indicates that increased human involvement in poverty estimation diminishes accuracy compared to machine learning involvement. Expert defined features exhibited significant overlap and poor interaction when used together in a classifier. Conversely, the CNN-based approach outperformed human experts, demonstrating superior predictive capability with low-resolution images. Significance : These findings highlight the importance of leveraging machine learning explainability methods to identify predictive elements that may be overlooked by human experts. This study advocates for the integration of emerging technologies with traditional methodologies to optimize data collection and analysis of poverty and welfare. Physical sciences/Mathematics and computing/Computer science Earth and environmental sciences/Environmental social sciences/Environmental economics Welfare estimation satellite imagery domain experts explainable AI convolutional neural networks Full Text Additional Declarations No competing interests reported. Supplementary Files supplimentarymaterial.pdf Cite Share Download PDF Status: Published Journal Publication published 02 Oct, 2024 Read the published version in Scientific Reports → Version 1 posted Editorial decision: Revision requested 23 Jul, 2024 Reviews received at journal 20 Jul, 2024 Reviewers agreed at journal 15 Jul, 2024 Reviews received at journal 14 Jul, 2024 Reviewers agreed at journal 24 Jun, 2024 Reviewers invited by journal 24 Jun, 2024 Editor assigned by journal 24 Jun, 2024 Editor invited by journal 13 Jun, 2024 Submission checks completed at journal 13 Jun, 2024 First submitted to journal 07 Jun, 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-4545272","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":319473577,"identity":"78755006-df31-489b-af9c-e36b9391cfdb","order_by":0,"name":"Hamid Sarmadi","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAA2ElEQVRIiWNgGAWjYBACAyA+wAPlSCRUHEggVcsZIrUwwLUwthGhxZz9dOKBNwx18ubTDj+88XDenTwG/sMH8Gqx7MndcHAOw2HDObfTjC0Stz0rZpBIw2+TwYHcDYd5GA4wzpBOMJNI3HY4sUGCxwC/lvNvQVrq7GdIp3+TSJwD1MJ//gN+LTfAtjAnzpDOAdrSANTCkINXB1DLW6BfDA4nA7UUWyQcO5zYJpFGyGG5mz+8qaizBTps480fNYcT+/kPP8BvDUQjEpuNCPWjYBSMglEwCggAAN7DT4uop8+1AAAAAElFTkSuQmCC","orcid":"","institution":"Halmstad University","correspondingAuthor":true,"prefix":"","firstName":"Hamid","middleName":"","lastName":"Sarmadi","suffix":""},{"id":319473580,"identity":"b638562a-9b94-40f0-b0b0-0e6f62bdc36c","order_by":1,"name":"Ibrahim Wahab","email":"","orcid":"","institution":"Lund University","correspondingAuthor":false,"prefix":"","firstName":"Ibrahim","middleName":"","lastName":"Wahab","suffix":""},{"id":319473581,"identity":"e7802334-7248-4de3-a3da-f3dd69ee06b9","order_by":2,"name":"Ola Hall","email":"","orcid":"","institution":"Lund University","correspondingAuthor":false,"prefix":"","firstName":"Ola","middleName":"","lastName":"Hall","suffix":""},{"id":319473582,"identity":"ce76f24e-33d4-44d9-a9c0-beb1b2b20ed2","order_by":3,"name":"Thorsteinn Rögnvaldsson","email":"","orcid":"","institution":"Halmstad University","correspondingAuthor":false,"prefix":"","firstName":"Thorsteinn","middleName":"","lastName":"Rögnvaldsson","suffix":""},{"id":319473583,"identity":"9275c176-71d9-42e8-b1d6-0308a7e82bc1","order_by":4,"name":"Mattias Ohlsson","email":"","orcid":"","institution":"Halmstad University","correspondingAuthor":false,"prefix":"","firstName":"Mattias","middleName":"","lastName":"Ohlsson","suffix":""}],"badges":[],"createdAt":"2024-06-07 09:49:35","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-4545272/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-4545272/v1","draftVersion":[],"editorialEvents":[{"content":"https://doi.org/10.1038/s41598-024-74150-9","type":"published","date":"2024-10-02T15:57:33+00:00"}],"editorialNote":"","failedWorkflow":false,"files":[{"id":66097123,"identity":"cd541514-dd01-49dc-8b7d-72cdc75e06f5","added_by":"auto","created_at":"2024-10-07 16:13:43","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1290205,"visible":true,"origin":"","legend":"","description":"","filename":"article.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4545272/v1_covered_3df66971-9713-4919-aea1-5c74f39323ae.pdf"},{"id":59238469,"identity":"dabf4f8e-6cfc-4a7a-b3b5-d47d43c4283f","added_by":"auto","created_at":"2024-06-28 04:46:03","extension":"pdf","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":4875731,"visible":true,"origin":"","legend":"","description":"","filename":"supplimentarymaterial.pdf","url":"https://assets-eu.researchsquare.com/files/rs-4545272/v1/86e77eca9d02c7637228f1d2.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Beyond Human Bias: Learning from CNNs’\r\nSuperior Insights in Satellite-Based Poverty\r\nMapping","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":"Welfare estimation, satellite imagery, domain experts, explainable AI, convolutional neural networks","lastPublishedDoi":"10.21203/rs.3.rs-4545272/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-4545272/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003e\u003cstrong\u003eBackground\u003c/strong\u003e: Satellite imagery is a potent tool for estimating human wealth and poverty, especially in regions lacking reliable data. This study compares a range of poverty estimation approaches from satellite images, spanning from expert-based to fully machine learning-based methodologies.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eMethods\u003c/strong\u003e: Human experts ranked clusters from the DHS survey using high-resolution satellite images. Then expert-defined features were utilized in a machine learning algorithm to estimate poverty. An explainability method was applied to assess the importance and interaction of these features in poverty prediction. Additionally, a convolutional neural network (CNN) was employed to estimate poverty from low-resolution satellite images of the same locations.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eFindings\u003c/strong\u003e: Our analysis indicates that increased human involvement in poverty estimation diminishes accuracy compared to machine learning involvement. Expert defined features exhibited significant overlap and poor interaction when used together in a classifier. Conversely, the CNN-based approach outperformed human experts, demonstrating superior predictive capability with low-resolution images.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eSignificance\u003c/strong\u003e: These findings highlight the importance of leveraging machine learning explainability methods to identify predictive elements that may be overlooked by human experts. This study advocates for the integration of emerging technologies with traditional methodologies to optimize data collection and analysis of poverty and welfare.\u003c/p\u003e","manuscriptTitle":"Beyond Human Bias: Learning from CNNs’\nSuperior Insights in Satellite-Based Poverty\nMapping","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2024-06-28 04:45:58","doi":"10.21203/rs.3.rs-4545272/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2024-07-23T05:37:58+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-20T08:22:55+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"111953809647229950310325940037243406298","date":"2024-07-15T22:19:23+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2024-07-14T23:12:52+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"115429679937797242444409734425758658013","date":"2024-06-24T20:36:46+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2024-06-24T16:14:05+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2024-06-24T16:10:51+00:00","index":"","fulltext":""},{"type":"editorInvited","content":"","date":"2024-06-13T09:29:05+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2024-06-13T09:18:09+00:00","index":"","fulltext":""},{"type":"submitted","content":"Scientific Reports","date":"2024-06-07T09:48:20+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":"ba429687-8d23-4e5a-b3a0-ec49d84fb591","owner":[],"postedDate":"June 28th, 2024","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"published-in-journal","subjectAreas":[{"id":33780205,"name":"Physical sciences/Mathematics and computing/Computer science"},{"id":33780206,"name":"Earth and environmental sciences/Environmental social sciences/Environmental economics"}],"tags":[],"updatedAt":"2024-10-07T16:07:03+00:00","versionOfRecord":{"articleIdentity":"rs-4545272","link":"https://doi.org/10.1038/s41598-024-74150-9","journal":{"identity":"scientific-reports","isVorOnly":false,"title":"Scientific Reports"},"publishedOn":"2024-10-02 15:57:33","publishedOnDateReadable":"October 2nd, 2024"},"versionCreatedAt":"2024-06-28 04:45:58","video":"","vorDoi":"10.1038/s41598-024-74150-9","vorDoiUrl":"https://doi.org/10.1038/s41598-024-74150-9","workflowStages":[]},"version":"v1","identity":"rs-4545272","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-4545272","identity":"rs-4545272","version":["v1"]},"buildId":"_2-kVJe1T_tPrBINL-cwx","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}