Poverty Mapping in the Age of Machine Learning

doi:10.1596/1813-9450-10429

Publication:
Poverty Mapping in the Age of Machine Learning

dc.contributor.author	Corral, Paul
dc.contributor.author	Segovia, Sandra
dc.date.accessioned	2023-05-04T13:53:48Z
dc.date.available	2023-05-04T13:53:48Z
dc.date.issued	2023-05-04
dc.description.abstract	Recent years have witnessed considerable methodological advances in poverty mapping, much of which has focused on the application of modern machine-learning approaches to remotely sensed data. Poverty maps produced with these methods generally share a common validation procedure, which assesses model performance by comparing subnational machine-learning-based poverty estimates with survey-based, direct estimates. Although unbiased, survey-based estimates at a granular level can be imprecise measures of true poverty rates, meaning that it is unclear whether the validation procedures used in machine-learning approaches are informative of actual model performance. This paper examines the credibility of existing approaches to model validation by constructing a pseudo-census from the Mexican Intercensal Survey of 2015, which is used to conduct several design-based simulation experiments. The findings show that the validation procedure often used for machine-learning approaches can be misleading in terms of model assessment since it yields incorrect information for choosing what may be the best set of estimates across different methods and scenarios. Using alternative validation methods, the paper shows that machine-learning-based estimates can rival traditional, more data intensive poverty mapping approaches. Further, the closest approximation to existing machine-learning approaches, using publicly available geo-referenced data, performs poorly when evaluated against “true” poverty rates and fails to outperform traditional poverty mapping methods in targeting simulations.	en
dc.identifier	http://documents.worldbank.org/curated/en/099759405012313710/IDU0c1878458051a40470808a960c8b70982671b
dc.identifier.doi	10.1596/1813-9450-10429
dc.identifier.uri	https://openknowledge.worldbank.org/handle/10986/39783
dc.language	English
dc.language.iso	en
dc.publisher	World Bank, Washington, DC
dc.relation.ispartofseries	Policy Research Working Papers; 10429
dc.rights	CC BY 3.0 IGO
dc.rights.holder	World Bank
dc.rights.uri	https://creativecommons.org/licenses/by/3.0/igo/
dc.subject	SMALL AREA ESTIMATION
dc.subject	POVERTY MAPPING
dc.subject	MACHINE LEARNING
dc.subject	SATELLITE IMAGERY
dc.subject	POVERTY RATE ESTIMATE
dc.subject	POVERTY DATA ANALYSIS
dc.title	Poverty Mapping in the Age of Machine Learning	en
dc.type	Working Paper
dspace.entity.type	Publication
okr.crossref.title	Poverty Mapping in the Age of Machine Learning
okr.date.disclosure	2023-05-01
okr.date.lastmodified	2023-05-01T00:00:00Z	en
okr.doctype	Policy Research Working Paper
okr.doctype	Publications & Research
okr.docurl	http://documents.worldbank.org/curated/en/099759405012313710/IDU0c1878458051a40470808a960c8b70982671b
okr.guid	099759405012313710
okr.identifier.docmid	IDU-c1878458-51a4-4708-8a96-c8b70982671b
okr.identifier.doi	10.1596/1813-9450-10429
okr.identifier.doi	http://dx.doi.org/10.1596/1813-9450-10429
okr.identifier.externaldocumentum	34053461
okr.identifier.internaldocumentum	34053461
okr.identifier.report	WPS10429
okr.import.id	597
okr.imported	true	en
okr.language.supported	en
okr.pdfurl	http://documents.worldbank.org/curated/en/099759405012313710/pdf/IDU0c1878458051a40470808a960c8b70982671b.pdf	en
okr.region.country	Mexico
okr.topic	International Economics and Trade::Economic Geography
okr.topic	Macroeconomics and Economic Growth::Economic Theory & Research
okr.topic	Science and Technology Development::Research and Development
okr.unit	EFI-AFR2-POV-Poverty and Equity (EAWPV)
relation.isAuthorOfPublication	562ed170-5271-5853-9c3c-a3ad9e16e176
relation.isAuthorOfPublication.latestForDiscovery	562ed170-5271-5853-9c3c-a3ad9e16e176