cerlymarco/MEDIUM_NoteBook

Covers practical implementations across time series forecasting, causal inference, anomaly detection, and model explainability using scikit-learn, gradient boosting, and SHAP. Demonstrates advanced techniques including conformal prediction intervals, transfer learning approaches, online learning, and drift detection using both traditional ML and modern interpretability tools. Each notebook pairs reproducible code with detailed methodological exploration of trade-offs in model selection, feature engineering, and handling imbalanced or missing data in production scenarios.

2,138 stars. No commits in the last 6 months.