telilagsr-9015-17-lab/child-undernutrition-ml-ethiopia

Machine learning and deep learning models for predicting stunting, wasting, and underweight among under-five children in Ethiopia using MCP survey data. # Child Undernutrition Prediction Using Machine Learning and Deep Learning This repository contains the full implementation of a research study on predicting stunting, wasting, and underweight among under-five children in Ethiopia using the Multidimensional Child Poverty (MCP) survey data. ## 📊 Objectives - Predict childhood undernutrition using ML and DL models - Compare multiple algorithms under different class imbalance strategies - Apply Explainable AI (SHAP) for model interpretation ## 🧠 Models Used - Logistic Regression - Random Forest - SVM - Gradient Boosting Machine (GBM) - XGBoost - MLP - TabNet ## ⚖️ Imbalance Handling - Class Weighting - SMOTEENN - No Balancing ## 🔍 Explainability - SHAP values - Feature importance analysis ## 👨‍💻 Author Telila Kejela(Msc in Data Science student at AAU) ## Supervisors - Dr Getachew Hailemariam(AAU) - Prof Abera Kumie(AAU) ## 📌 Note This project is part of a postgraduate research thesis in AAU, Ethiopia.