Healthcare demand forecasting and staffing decision platform (PoC): a 13-model benchmark (baselines, SARIMAX, Prophet, global gradient boosting, Nixtla Stats/ML/Neural, Chronos), conformal prediction intervals, rolling-origin backtesting, a costed staffing layer, batch pipeline, FastAPI serving, monitoring and responsible-ML docs.
python data-science machine-learning time-series scikit-learn healthcare xgboost forecasting lightgbm uncertainty-quantification prophet sarimax portfolio-project time-series-forecasting conformal-prediction mlops fastapi nixtla staffing-optimization model-benchmark
-
Updated
Jun 20, 2026 - Jupyter Notebook