TF Boosted Trees: A scalable TensorFlow based framework for gradient boosting
This provides a new tool for machine learning practitioners needing efficient distributed training of gradient boosted trees, though it appears incremental as it builds on existing gradient boosting methods.
The paper introduces TF Boosted Trees (TFBT), a scalable TensorFlow-based framework for distributed gradient boosting, featuring a novel architecture and techniques like layer-by-layer boosting that achieve smaller ensembles and faster prediction.
TF Boosted Trees (TFBT) is a new open-sourced frame-work for the distributed training of gradient boosted trees. It is based on TensorFlow, and its distinguishing features include a novel architecture, automatic loss differentiation, layer-by-layer boosting that results in smaller ensembles and faster prediction, principled multi-class handling, and a number of regularization techniques to prevent overfitting.