JoyNgaru/Kenya-Food-Security-Risk-Clustering-Using-Kmeans-and-DBScans

# Kenya Food Security Risk Clustering Using K-means and DBScans About Dataset This dataset supports the development of a machine learning early-warning system for food insecurity across Kenya's Arid and Semi-Arid Lands (ASAL). Kenya Food Security Risk It combines three critical data streams — IPC food insecurity outcomes, CHIRPS rainfall indicators, and MODIS NDVI vegetation features — into a single county-level, time-aligned resource for exploratory analysis, predictive modeling, and humanitarian research. Current project stage: Data preparation, feature engineering, and exploratory analysis. Next stage: Baseline classification model to predict high-risk food insecurity cases 1–3 months in advance. The Problem Recent IPC projections indicate that millions of Kenyans may face IPC Phase 3 "Crisis" level food insecurity during severe drought periods. Kenya's ASAL counties are particularly vulnerable to drought-driven food crises. Humanitarian agencies often react to crises after they peak, not before. There is limited access to simple, county-level early-warning tools that connect food insecurity outcomes with publicly available environmental indicators such as rainfall and vegetation health. This dataset was built to help close that gap by providing a clean, merged resource that links environmental stress to food insecurity severity at the county level. Data Sources Dataset Source Type Coverage IPC Acute Food Insecurity FEWS NET / HDX Food insecurity phase classifications (Phase 1–5) and population percentages Kenya ASAL counties CHIRPS Rainfall UCSB Climate Hazards Center Monthly rainfall estimates with station calibration 2019–2026 MODIS NDVI NASA/USGS via Google Earth Engine Satellite vegetation health indices from MODIS Terra 2019–2026 Kenya County Boundaries HDX / administrative boundaries County shapefiles and GeoJSON for spatial alignment 23 ASAL counties Files in This Dataset 1. ipc_rainfall_ndvi_master_dataset.cs …