TF Boosted Trees：一个基于 TensorFlow 的可扩展梯度提升框架
机器学习
2017-11-01 v1 机器学习
摘要
TF Boosted Trees (TFBT) 是一个用于梯度提升树分布式训练的新的开源框架。它基于 TensorFlow，其显著特性包括一种新颖的架构、自动损失求导、逐层提升（从而产生更小的集成与更快的预测）、原则性的多类处理，以及一系列用于防止过拟合的正则化技术。
关键词
引用
@article{arxiv.1710.11555,
title = {TF Boosted Trees: A scalable TensorFlow based framework for gradient boosting},
author = {Natalia Ponomareva and Soroush Radpour and Gilbert Hendry and Salem Haykal and Thomas Colthurst and Petr Mitrichev and Alexander Grushetsky},
journal= {arXiv preprint arXiv:1710.11555},
year = {2017}
}
备注
European Conference on Machine Learning and Principles and Practice of Knowledge Discovery in Databases (ECML PKDD 2017). The final publication will be available at link.springer.com and is available on ECML website http://ecmlpkdd2017.ijs.si/papers/paperID705.pdf