TF Boosted Trees: A Scalable TensorFlow Based Framework for Gradient Boosting

Natalia Ponomareva, Soroush Radpour, Gilbert Hendry, Salem Haykal, Thomas Colthurst, Petr Mitrichev, Alexander Grushetsky

ECML-PKDD 2017 pp. 423-427

doi:10.1007/978-3-319-71273-4_44 /ecmlpkdd/2017/ponomareva2017ecmlpkdd-tf/

Abstract

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.

PDF ECML-PKDD Semantic Scholar

Cite

Text

Ponomareva et al. "TF Boosted Trees: A Scalable TensorFlow Based Framework for Gradient Boosting." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2017. doi:10.1007/978-3-319-71273-4_44

Markdown

[Ponomareva et al. "TF Boosted Trees: A Scalable TensorFlow Based Framework for Gradient Boosting." European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases, 2017.](https://mlanthology.org/ecmlpkdd/2017/ponomareva2017ecmlpkdd-tf/) doi:10.1007/978-3-319-71273-4_44

BibTeX

@inproceedings{ponomareva2017ecmlpkdd-tf,
  title     = {{TF Boosted Trees: A Scalable TensorFlow Based Framework for Gradient Boosting}},
  author    = {Ponomareva, Natalia and Radpour, Soroush and Hendry, Gilbert and Haykal, Salem and Colthurst, Thomas and Mitrichev, Petr and Grushetsky, Alexander},
  booktitle = {European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases},
  year      = {2017},
  pages     = {423-427},
  doi       = {10.1007/978-3-319-71273-4_44},
  url       = {https://mlanthology.org/ecmlpkdd/2017/ponomareva2017ecmlpkdd-tf/}
}