AI-Powered Poverty Mapping: New Frontiers in Targeting Social Protection

D4Act's data science team has developed machine learning models that predict household-level poverty with over 80% accuracy using freely available data sources - satellite nighttime light imagery, daytime land use classification, mobile phone call detail records, and publicly available geographic features. These models enable governments to target social protection programmes more precisely, reaching the poorest households while reducing both inclusion and exclusion errors.