The library provides gradient boosted trees classification and regression algorithms based on an ensemble of regression (decision) trees trained using stochastic gradient boosting technique.

Regression tree

is a binary tree graph. Its internal (split) nodes represent a

decision function

used to select following (child) node at prediction stage. Its leaf (terminal) nodes represent the corresponding response values which are the result of prediction from the tree. For more details, see Decision Tree [Breiman84].