Scalable and Flexible Gradient Boosting

Latest from the XGBoost Blog

Supports regression, classification, ranking, and user-defined objectives.

Runs on Windows, Linux, and macOS, as well as major cloud platforms.

Supports multiple languages including C++, Python, R, Java, Scala, and Julia.

Wins many data science and machine learning challenges and is used in production by multiple companies.

Supports distributed training across multiple machines and integrates with systems including Spark and Dask.

Its optimized backend delivers strong performance with limited resources and scales beyond billions of examples.