XGBoost Achieves Accurate AMR Trend Forecasts

Researchers present a two-component framework (Feb 26, 2026) to forecast antimicrobial resistance using WHO GLASS data from 2021–2023, benchmarking six models across 5,909 observations and six WHO regions. XGBoost performed best (test MAE 7.07%, R² 0.854), prior-year resistance dominated feature importance (50.5%), regional MAE ranged 4.16%–10.14%, and a RAG pipeline with ChromaDB and Phi-3 Mini supports source-attributed policy answers.