PMML Gradient Boosted Trees Predictor

Applies classification from a Gradient Boosted Trees model that is provided in PMML format. Note that it is currently not possible to load models that were learned on a bit-, byte- or double-vector column and then written to PMML because PMML does not support vector columns.

The implementation follows the algorithms described in Greedy Function Approximation: A Gradient Boosting Machine by Jerome H. Friedman (1999).

Input ports

1. Type: PMML
   Model
   Gradient Boosted Trees model in PMML format.
2. Type: Table
   Input Data
   The data to predict.

Output ports

1. Type: Table
   Output Data
   The predicted data.

External resources

• Wikipedia: Gradient Boosting