Catholic Relief Services Leverages Machine Learning to Fight Hunger

Summary

Feature article on CRS MIRA system. Details the embedded enumerator methodology, data collection approach (demographics, livelihood, shock history, coping strategies), UBALE programme context (five-year, USD 63 million, 250,000 households, 284 communities), use of KNN and LASSO algorithms, Dartmouth Flood Observatory maps for household selection, operational sharing of early warning data with village development committees, Microsoft AI for Humanitarian Action Grant, FutureEdge 50 Award, and planned expansion to plug-and-play solution.