Oselin1988/LRD_STUDY

Machine learning analysis of lower respiratory infection mortality in 16 West African countries (2010–2023). Implements k-means clustering, XGBoost with SHAP, immune arm transition analysis, COVID-19 counterfactual, and ensemble forecasting (Prophet, XGBoost, LightGBM, LSTM) to 2030. # LRI Mortality in West Africa: Machine Learning Analysis ## Overview This repository contains the complete code and data for the study: *"Temporal Trends and Mortality Burden of Lower Respiratory Infections in West African Countries (2010–2023): A Global Burden of Disease Analysis"* The study integrates: - Unsupervised clustering of mortality trajectories (k‑means) - Machine learning prediction (XGBoost) with SHAP interpretation - Temporal SHAP and immune arm transition analysis - COVID‑19 counterfactual and PM₂.₅ reduction scenarios - Ensemble forecasting (Prophet, XGBoost, LightGBM, LSTM) to 2030 All analyses are fully reproducible. ## Repository structure west-africa-lri-ml/ ├── data/ # Cleaned dataset (CSV) ├── scripts/ # Analysis scripts (01–06) ├── results/ # Output CSV tables ├── figures/ # Publication‑ready figures ├── requirements.txt # Python dependencies ├── LICENSE # MIT license └── README.md # This file