Using machine learning to produce a high-resolution poverty map of sub-Saharan Africa

Fighting poverty requires knowing where the poor live. As such, poverty maps represent an essential tool for designing, targeting, and monitoring poverty reduction policies. Traditionally poverty maps have been constructed based on national household surveys and census data. However, traditional poverty maps are often not available for developing countries or, when they are, they have a limited spatial resolution or are outdated. What can be done in these cases? New methods based on artificial intelligence and non-traditional sources of data are rapidly expanding the possibilities to produce detailed poverty maps also in data-scarce settings.

In this webinar, the second of a series on the use of artificial intelligence for poverty analysis, Kamwoo Lee will present a new method to map poverty at the village level using machine learning and readily available geospatial data including OpenStreetMap and satellite images. The method has been applied to produce a high-resolution poverty map of Sub-Saharan Africa, including for countries lacking survey and census data. In the webinar, Dr. Lee will discuss the accuracy of the new method and its implication for real-world policymaking.

SPEAKER:

Kamwoo Lee is a Data Scientist at the World Bank. He works in the Data Analytics and Tools Unit (DECAT), where he is currently developing a data platform that facilitates the discovery and dissemination of datasets related to fragile, conflict, and violence-affected (FCV) countries. He received a Ph.D. in Systems and Information Engineering from the University of Virginia in 2020. His academic research focused on applying machine learning and simulation techniques to various global development projects. His previous project topics include social conflict and segregation, financial inclusion and cryptocurrency, and public healthcare and traditional healing. Before his Ph.D. study, he worked as a financial regulator for 6 years in the Financial Supervisory Service of Korea and an R&D engineer at Samsung Electronics as well as an international volunteer in the Philippines.

Webcast