Tutorials

Learn boosters through hands-on Jupyter notebook tutorials. Each tutorial is self-contained and can be run independently.

Difficulty Legend

• 🟢 Beginner — No prior boosting experience needed

• 🟡 Intermediate — Familiarity with ML concepts helpful

Tutorials

🟢 01. Basic GBDT Training

Train your first GBDT model from scratch.

Tutorial 01: Basic GBDT Training

🟢 02. sklearn Integration

Use boosters with sklearn pipelines and cross-validation.

Tutorial 02: sklearn Integration

🟢 03. Binary Classification

Train a classifier with AUC and ROC curves.

Tutorial 03: Binary Classification

🟡 04. Multiclass Classification

Handle multiple classes with softmax objective.

Tutorial 04: Multiclass Classification

🟡 05. Early Stopping

Prevent overfitting with validation-based stopping.

Tutorial 05: Early Stopping

🟡 06. GBLinear & Sparse Data

Use linear boosting for high-dimensional sparse data.

Tutorial 06: GBLinear & Sparse Data

🟡 07. Hyperparameter Tuning

Find optimal hyperparameters for your data.

Tutorial 07: Hyperparameter Tuning

🟡 08. Explainability

Interpret models with feature importance and SHAP.

Tutorial 08: Explainability

🟡 09. Model Serialization

Save, load, and convert models between formats.

Tutorial 09: Model Serialization

🟡 10. Linear Trees

GBDT with linear regression in leaves.

Tutorial 10: GBDT with Linear Leaves

🟡 11. Sample Weighting

Handle class imbalance and concept drift.

Tutorial 11: Sample Weighting