christianadebambo/uganda-hwc-policy

# Uncertainty-Aware Decision Support for Human-Wildlife Conflict in Uganda ## Overview We develop a reproducible pipeline for human–wildlife conflict (HWC) decision support in Uganda’s Kasese District, combining: - Severity modelling (logistic regression, TabTransformer) - Calibration and uncertainty (Platt scaling, temperature scaling, conformal prediction) - Uplift modelling (multi-arm T-learners with XGBoost) - Off-policy evaluation (IPS, overlap weighting, doubly robust) ## Data Dataset: _kasese-hwc-data-2021-combined-2021-2022-partly-cleaned.csv_ ## Installation Clone this repository: ```bash git clone github.com cd uganda-hwc-policy ``` Create a virtual environment and install dependencies: ```bash python -m venv .venv source .venv/bin/activate pip install -r requirements.txt ``` ## Usage Open the notebook: ```bash jupyter notebook uganda-hwc-policy.ipynb ``` The notebook runs end-to-end: - Loads raw CSV (update path if running locally) - Cleans and engineers features - Trains models - Saves outputs under _outputs/_ _**N.B: Adjust file paths accordingly**_ ## Outputs Key artefacts saved in _outputs/_: - _reliability_valid.png_ - calibration diagram - _logistic_top_weights.csv_ - logistic regression weights - _tabtransformer_perm_importance.csv_ - feature importance - _uplift_recommendations_by_parish.csv_ - parish-level recommendations