About This Encyclopedia

Editorial Stance

The XGBoost Reader's Encyclopedia is an independent, community-maintained knowledge resource about the XGBoost gradient boosting algorithm. It is not a replacement for the official XGBoost documentation at xgboost.readthedocs.io.

Our purpose is complementary: we provide concept explanations for those who want to understand the "why" behind the "how", mathematical derivations with explicit steps (not just final formulas), annotated readings of key papers (Chen & Guestrin 2016, the original), and unbiased benchmarks against LightGBM, CatBoost, and other gradient boosting implementations.

We are independent contributors with no affiliation to the official XGBoost project, DMLC, or any vendor. We publish this encyclopedia because we love the algorithm and believe its inner workings deserve clear exposition.

Methodology

• All mathematical derivations are checked against the original papers and (where possible) against the XGBoost source code.
• Benchmarks are run on controlled hardware with published parameters — you can reproduce them.
• Code examples are Python-based, using the official xgboost package, tested against the latest stable release.