XGBoost

1. Introduction to boosted trees

   Gradient boosted trees, objective, regularization, and the additive training model.

2. Python package introduction

   DMatrix, training, prediction, and saving models.

3. Parameters

   General, booster, and task parameters, including eta, max_depth, and objectives.

4. Notes on parameter tuning

   Bias-variance tradeoff, overfitting control, and imbalanced data.

5. Scikit-learn interface

   XGBClassifier and XGBRegressor with scikit-learn tools.

6. Categorical data

   Native categorical feature support.

7. Custom objective and metric

   Writing custom loss functions and evaluation metrics.

8. Model IO

   Saving and loading models in JSON and UBJSON formats.

9. GPU support

   Training and prediction on CUDA GPUs.

10. Distributed XGBoost

    Dask, Spark, and other distributed training options.

11. Monotonic and interaction constraints

    Constraining model structure for interpretability.

12. External memory

    Training on data larger than memory.

13. Learning to rank

    Ranking objectives for search and recommendation.

Each section links to the official page. Titles and summaries by Extra Credit; the docs belong to their authors.