Trains gradient-boosted tree models across local and distributed environments, with bindings for Python, R, JVM, Julia, and C++. Its sparsity-aware split finding and quantile sketch made it a default baseline for tabular ML competitions.
Trains gradient-boosted decision trees with native categorical-feature handling, GPU acceleration, and production-ready prediction APIs. A strong fit for tabular ML when preprocessing categories into numeric features would add noise or leakage.