CatBoost

CatBoost (Yandex) is arguably the easiest boosting library to use because it handles Categorical Features automatically and perfectly without tuning.

When to Use

• Categorical Data: If you have many strings/IDs, CatBoost is king.
• Default Params: Works incredibly well out of the box.

Core Concepts

Ordered Boosting

A technique to avoid target leakage (overfitting) during training.

Symmetric Trees

Builds balanced trees, which are faster at inference time.

Best Practices (2025)

Do:

• Use pool: Pool() is efficient for data loading.
• Use GPU: CatBoost's GPU implementation is highly optimized.

Don't:

• Don't One-Hot Encode: Let CatBoost handle it natively.

References

• CatBoost Documentation