Development version

You are reading the latest (development) docs. This version tracks the master branch and may contain unreleased features or breaking changes. For stable documentation, see stable.

Tutorials

Step-by-step tutorials for common real-world scenarios. Each tutorial is self-contained and includes a baseline comparison, runnable code, visualizations, and practical notes.

Tutorials vs Examples

Tutorials are end-to-end walkthroughs of a complete real-world task — from raw data to a tuned, evaluated model — usually integrating a specific library (XGBoost, LightGBM, CatBoost, …). Examples are shorter, focused recipes that each demonstrate a single feature you can drop into your own code.

Gradient Boosting Libraries

Tutorial	What it covers
Tune XGBoost	9-parameter XGBoost search, adaptive schedules, feature importance, 3-way comparison
Tune LightGBM	9-parameter LightGBM search, num_leaves/max_depth interaction, parameter scatter plots
Tune CatBoost	7-parameter CatBoost search, bagging_temperature, border_count, GPU tip

Feature Selection

Tutorial	What it covers
Comprehensive Feature Selection	3-stage workflow: select on 50 features, retune on selected subset, validate with a second estimator

Imbalanced Data

Tutorial	What it covers
Imbalanced Classification	95/5 imbalance, class_weight as search param, balanced_accuracy scoring, confusion matrices

Outlier Detection

Tutorial	What it covers
Isolation Forest	Custom scorer from score_samples, 4-param search, anomaly contour plots, ROC curve

See Also

• Examples — shorter end-to-end examples for common use cases
• User Guide — decision guide for choosing a search method
• API Reference — full parameter documentation