面向高维特征和多分类的分布式梯度提升树
作者:
作者单位:

作者简介:

江佳伟(1990-),男,湖北洪湖人,博士,CCF学生会员,主要研究领域为机器学习;邵蓥侠(1988-),男,博士,副研究员,博士生导师,CCF专业会员,主要研究领域为数据库,知识图谱数据管理,并行图计算,知识工程;符芳诚(1996-),男,学士,主要研究领域为机器学习;崔斌(1975-),男,博士,教授,博士生导师,CCF杰出会员,主要研究领域为数据库,大数据管理分析.

通讯作者:

崔斌,E-mail:bin.cui@pku.edu.cn

中图分类号:

基金项目:

国家自然科学基金(61832001,61702015,61702016);国家重点研发计划(2018YFB1004403)


Distributed Gradient Boosting Decision Tree Algorithm for High-dimensional and Multi-classification Problems
Author:
Affiliation:

Fund Project:

National Natural Science Foundation of China (61702423, 61532021, U1501252, 61402180); National Key Research and Development Program of China (2016YFB1000905)

  • 摘要
  • |
  • 图/表
  • |
  • 访问统计
  • |
  • 参考文献
  • |
  • 相似文献
  • |
  • 引证文献
  • |
  • 资源附件
  • |
  • 文章评论
    摘要:

    梯度提升树算法由于其高准确率和可解释性,被广泛地应用于分类、回归、排序等各类问题.随着数据规模的爆炸式增长,分布式梯度提升树算法成为研究热点.虽然目前已有一系列分布式梯度提升树算法的实现,但是它们在高维特征和多分类任务上性能较差,原因是它们采用的数据并行策略需要传输梯度直方图,而高维特征和多分类情况下梯度直方图的传输成为性能瓶颈.针对这个问题,研究更加适合高维特征和多分类的梯度提升树的并行策略,具有重要的意义和价值.首先比较了数据并行与特征并行策略,从理论上证明特征并行更加适合高维和多分类场景.根据理论分析的结果,提出了一种特征并行的分布式梯度提升树算法FP-GBDT.FP-GBDT设计了一种高效的分布式数据集转置算法,将原本按行切分的数据集转换为按列切分的数据表征;在建立梯度直方图时,FP-GBDT使用一种稀疏感知的方法来加快梯度直方图的建立;在分裂树节点时,FP-GBDT设计了一种比特图压缩的方法来传输数据样本的位置信息,从而减少通信开销.通过详尽的实验,对比了不同并行策略下分布式梯度提升树算法的性能,首先验证了FP-GBDT提出的多种优化方法的有效性;然后比较了FP-GBDT与XGBoost的性能,在多个数据集上验证了FP-GBDT在高维特征和多分类场景下的有效性,取得了最高6倍的性能提升.

    Abstract:

    Gradient boosting decision tree algorithm is widely used in various tasks, such as classification, regression, and ranking, owing to its high accuracy and strong interpretability. With the explosive growth of data volume, distributed gradient boosting decision tree algorithms have become an important research issue. Although there exists a series of implementations of distributed gradient boosting decision tree, they perform poorly on high-dimensional and multi-classification tasks. The data parallel strategy they adopt requires the transmission of gradient histograms, and this communication overhead becomes the bottleneck in many high-dimensional and multi-classification task. This study aims at this problem and tries to find an efficient parallel strategy that is more suitable for the target. Data-parallel and feature-parallel strategies are first compared based on a cost model, and it is theoretically proved that feature-parallel is more suitable for high-dimensional and multi-classification tasks. Based on the analysis, this paper proposes a feature-parallel distributed gradient boosting decision tree algorithm, named FP-GBDT. FP-GBDT designs an efficient distributed dataset transposition method to partition the training dataset by column. During the construction of gradient histogram, FP-GBDT uses a sparsity-aware method to accelerate the histogram construction. When splitting tree nodes, FP-GBDT develops a bitmap compression method to transmit the placement of instances, thereby reduces the communication overhead. This study compares the performance of distributed gradient boosting decision tree algorithm under different parallel strategies through extensive experiments. First, the effectiveness of proposed optimization methods in FP-GBDT is verified. Then, the representative of data-parallel strategy of FP-GBDT and XGBoost are compared. On various datasets, it is proved that FP-GBDT is more efficient in high-dimensional and multi-classification tasks. FP-GBDT achieves up to 6 times performance improvement than data-parallel implementations.

    参考文献
    相似文献
    引证文献
引用本文

江佳伟,符芳诚,邵蓥侠,崔斌.面向高维特征和多分类的分布式梯度提升树.软件学报,2019,30(3):784-798

复制
分享
文章指标
  • 点击次数:
  • 下载次数:
  • HTML阅读次数:
  • 引用次数:
历史
  • 收稿日期:2018-07-19
  • 最后修改日期:2018-09-20
  • 录用日期:
  • 在线发布日期: 2019-03-06
  • 出版日期:
文章二维码
您是第位访问者
版权所有:中国科学院软件研究所 京ICP备05046678号-3
地址:北京市海淀区中关村南四街4号,邮政编码:100190
电话:010-62562563 传真:010-62562533 Email:jos@iscas.ac.cn
技术支持:北京勤云科技发展有限公司

京公网安备 11040202500063号