getnetbogale27/explainable-ml-child-undernutrition

Reproducible Python implementation of explainable, imbalance-aware machine-learning models for classifying concurrent child undernutrition outcomes using Ethiopian Young Lives Cohort Study data. # Imbalance-Aware Explainable Machine Learning for Child Undernutrition Modeling This repository contains reproducible research code and scripts for the manuscript "Imbalance-Aware Explainable Machine Learning for Concurrent Child Undernutrition Modeling" (Young Lives Ethiopia). It is structured to enable end-to-end regeneration of preprocessing, model training, evaluation, tables, and figures without shipping restricted data. ## Repository map - **Configuration** - `configs/` — Model, experiment, or pipeline configuration files. - **Data** - `data/` — Raw and processed datasets (restricted data not included). - **Documentation** - `docs/` — Project documentation and references. - **Figures** - `figures/` — Generated plots and supplementary visual assets. - **Notebooks** - `notebooks/` — Exploratory analysis, prototyping, and reports. - **Scripts** - `scripts/` — CLI utilities and one-off automation tasks. - **Source code** - `src/` — Core library code for data prep, modeling, and explainability. - **Tests** - `tests/` — Automated tests for the codebase. - **Dependencies** - `requirements.txt` — Python dependency list. ## Quickstart 1. Create an environment: ```bash python -m venv .venv source .venv/bin/activate pip install -r requirements.txt ``` 2. Add data: - Place the dataset at `data/Baseline Data.csv` (or update `configs/default.txt`). - The dataset should contain the outcome column and covariates described in the manuscript. 3. Run the full pipeline: ```bash python scripts/run_pipeline.py --config configs/default.txt ``` 4. Reproduce figures and tables: ```bash python scripts/make_figures.py --config configs/default.txt python scripts/make_tables.py --config configs/default.txt ``` ## Supplementary tables (S5-S7) Generate specific supplementary tables directly from the raw dataset: ```bash python scripts/make_table_s5_bivariate.py --data data/young_lives_ethiopia.csv --target concurrent_conditions --out results/tables/table_s5_bivariate.csv python scr …