结合调查和普查数据,利用半监督深度学习改进贫困预测

IF 5.1 1区 经济学 Q1 ECONOMICS Journal of Development Economics Pub Date : 2024-10-10 DOI:10.1016/j.jdeveco.2024.103385
Damien Echevin , Guy Fotso , Yacine Bouroubi , Harold Coulombe , Qing Li
{"title":"结合调查和普查数据,利用半监督深度学习改进贫困预测","authors":"Damien Echevin ,&nbsp;Guy Fotso ,&nbsp;Yacine Bouroubi ,&nbsp;Harold Coulombe ,&nbsp;Qing Li","doi":"10.1016/j.jdeveco.2024.103385","DOIUrl":null,"url":null,"abstract":"<div><div>This paper presents a methodology for predicting poverty using semi-supervised learning techniques, specifically pseudo-labeling, and deep learning algorithms. Standard poverty prediction models rely on limited household survey data, whereas our approach exploits large amounts of unlabeled census data to improve prediction accuracy. By applying pseudo-labeling, we improve key performance metrics across various African regions, where our models outperform conventional approaches to identifying poor individuals. Deep neural networks (DNNs) trained on pseudo-labeled data exhibited area under the curve (AUC) scores ranging from 0.8 to over 0.9, a notable improvement over previous machine learning survey-based methods. Furthermore, random undersampling was key to refining model performance, balancing higher coverage with some reduction in precision. These findings have significant implications for poverty targeting, enabling more accurate identification of poor individuals and supporting better resource allocation.</div></div>","PeriodicalId":48418,"journal":{"name":"Journal of Development Economics","volume":"172 ","pages":"Article 103385"},"PeriodicalIF":5.1000,"publicationDate":"2024-10-10","publicationTypes":"Journal Article","fieldsOfStudy":null,"isOpenAccess":false,"openAccessPdf":"","citationCount":"0","resultStr":"{\"title\":\"Combining survey and census data for improved poverty prediction using semi-supervised deep learning\",\"authors\":\"Damien Echevin ,&nbsp;Guy Fotso ,&nbsp;Yacine Bouroubi ,&nbsp;Harold Coulombe ,&nbsp;Qing Li\",\"doi\":\"10.1016/j.jdeveco.2024.103385\",\"DOIUrl\":null,\"url\":null,\"abstract\":\"<div><div>This paper presents a methodology for predicting poverty using semi-supervised learning techniques, specifically pseudo-labeling, and deep learning algorithms. Standard poverty prediction models rely on limited household survey data, whereas our approach exploits large amounts of unlabeled census data to improve prediction accuracy. By applying pseudo-labeling, we improve key performance metrics across various African regions, where our models outperform conventional approaches to identifying poor individuals. Deep neural networks (DNNs) trained on pseudo-labeled data exhibited area under the curve (AUC) scores ranging from 0.8 to over 0.9, a notable improvement over previous machine learning survey-based methods. Furthermore, random undersampling was key to refining model performance, balancing higher coverage with some reduction in precision. These findings have significant implications for poverty targeting, enabling more accurate identification of poor individuals and supporting better resource allocation.</div></div>\",\"PeriodicalId\":48418,\"journal\":{\"name\":\"Journal of Development Economics\",\"volume\":\"172 \",\"pages\":\"Article 103385\"},\"PeriodicalIF\":5.1000,\"publicationDate\":\"2024-10-10\",\"publicationTypes\":\"Journal Article\",\"fieldsOfStudy\":null,\"isOpenAccess\":false,\"openAccessPdf\":\"\",\"citationCount\":\"0\",\"resultStr\":null,\"platform\":\"Semanticscholar\",\"paperid\":null,\"PeriodicalName\":\"Journal of Development Economics\",\"FirstCategoryId\":\"96\",\"ListUrlMain\":\"https://www.sciencedirect.com/science/article/pii/S0304387824001342\",\"RegionNum\":1,\"RegionCategory\":\"经济学\",\"ArticlePicture\":[],\"TitleCN\":null,\"AbstractTextCN\":null,\"PMCID\":null,\"EPubDate\":\"\",\"PubModel\":\"\",\"JCR\":\"Q1\",\"JCRName\":\"ECONOMICS\",\"Score\":null,\"Total\":0}","platform":"Semanticscholar","paperid":null,"PeriodicalName":"Journal of Development Economics","FirstCategoryId":"96","ListUrlMain":"https://www.sciencedirect.com/science/article/pii/S0304387824001342","RegionNum":1,"RegionCategory":"经济学","ArticlePicture":[],"TitleCN":null,"AbstractTextCN":null,"PMCID":null,"EPubDate":"","PubModel":"","JCR":"Q1","JCRName":"ECONOMICS","Score":null,"Total":0}
引用次数: 0

摘要

本文介绍了一种利用半监督学习技术(特别是伪标签技术)和深度学习算法预测贫困的方法。标准的贫困预测模型依赖于有限的家庭调查数据,而我们的方法则利用大量未标记的人口普查数据来提高预测的准确性。通过应用伪标签技术,我们改进了非洲各地区的关键性能指标,在这些地区,我们的模型在识别贫困人口方面优于传统方法。在伪标签数据上训练的深度神经网络(DNN)的曲线下面积(AUC)得分从 0.8 到 0.9 以上不等,与之前基于调查的机器学习方法相比有显著提高。此外,随机欠采样是改进模型性能的关键,在提高覆盖率的同时也降低了精度。这些发现对确定贫困目标具有重要意义,可以更准确地识别贫困人口,支持更好的资源分配。
本文章由计算机程序翻译,如有差异,请以英文原文为准。
查看原文
分享 分享
微信好友 朋友圈 QQ好友 复制链接
本刊更多论文
Combining survey and census data for improved poverty prediction using semi-supervised deep learning
This paper presents a methodology for predicting poverty using semi-supervised learning techniques, specifically pseudo-labeling, and deep learning algorithms. Standard poverty prediction models rely on limited household survey data, whereas our approach exploits large amounts of unlabeled census data to improve prediction accuracy. By applying pseudo-labeling, we improve key performance metrics across various African regions, where our models outperform conventional approaches to identifying poor individuals. Deep neural networks (DNNs) trained on pseudo-labeled data exhibited area under the curve (AUC) scores ranging from 0.8 to over 0.9, a notable improvement over previous machine learning survey-based methods. Furthermore, random undersampling was key to refining model performance, balancing higher coverage with some reduction in precision. These findings have significant implications for poverty targeting, enabling more accurate identification of poor individuals and supporting better resource allocation.
求助全文
通过发布文献求助,成功后即可免费获取论文全文。 去求助
来源期刊
CiteScore
8.30
自引率
4.00%
发文量
126
审稿时长
72 days
期刊介绍: The Journal of Development Economics publishes papers relating to all aspects of economic development - from immediate policy concerns to structural problems of underdevelopment. The emphasis is on quantitative or analytical work, which is relevant as well as intellectually stimulating.
期刊最新文献
Editorial Board On the properties of the two main types of global poverty lines The local human capital costs of oil exploitation Cover more for less: Targeted drug coverage, chronic disease management, and medical spending Corrigendum to “Rural road stimulus and the role of matching mandates on economic recovery in China” [J. Dev. Econ. 166, (January 2024), 103211]
×
引用
GB/T 7714-2015
复制
MLA
复制
APA
复制
导出至
BibTeX EndNote RefMan NoteFirst NoteExpress
×
×
提示
您的信息不完整,为了账户安全,请先补充。
现在去补充
×
提示
您因"违规操作"
具体请查看互助需知
我知道了
×
提示
现在去查看 取消
×
提示
确定
0
微信
客服QQ
Book学术公众号 扫码关注我们
反馈
×
意见反馈
请填写您的意见或建议
请填写您的手机或邮箱
已复制链接
已复制链接
快去分享给好友吧!
我知道了
×
扫码分享
扫码分享
Book学术官方微信
Book学术文献互助
Book学术文献互助群
群 号:481959085
Book学术
文献互助 智能选刊 最新文献 互助须知 联系我们:info@booksci.cn
Book学术提供免费学术资源搜索服务,方便国内外学者检索中英文文献。致力于提供最便捷和优质的服务体验。
Copyright © 2023 Book学术 All rights reserved.
ghs 京公网安备 11010802042870号 京ICP备2023020795号-1