Nonparametric feature selection by random forests and deep neural networks

Nonparametric feature selection by random forests and deep neural networks
复制标题

通过随机森林和深度神经网络进行非参数特征选择

DOI:
10.1016/j.csda.2022.107436
复制
发表时间:
2022-01
影响因子:
1.8
通讯作者:
Zhonglei Wang
Zhonglei Wang
中科院分区:
数学3区
文献类型:
--
作者:
Xiaojun Mao;Liuhua Peng;Zhonglei Wang

文献摘要

参考文献

相似文献

随机森林是一种广泛使用的机器学习算法,但当应用于包含大量实例和无用特征的大规模数据集时,其计算效率会受到影响。本文提出了一种结合随机森林和深度神经网络的非参数特征选择算法,并研究了其在正则性条件下的理论性质。通过使用不同的合成模型和一个真实世界的例子,我们证明了该算法在识别有用特征、避免无用特征和计算效率方面的优势。虽然该算法是使用标准随机森林提出的,但它可以广泛适用于其他机器学习算法,只要特征可以进行相应的排序。
Random forests are a widely used machine learning algorithm, but their computational efficiency is undermined when applied to large-scale datasets with numerous instances and useless features. Herein, we propose a nonparametric feature selection algorithm that incorporates random forests and deep neural networks, and its theoretical properties are also investigated under regularity conditions. Using different synthetic models and a real-world example, we demonstrate the advantage of the proposed algorithm over other alternatives in terms of identifying useful features, avoiding useless ones, and the computation efficiency. Although the algorithm is proposed using standard random forests, it can be widely adapted to other machine learning algorithms, as long as features can be sorted accordingly.
DOI: 10.1016/j.cell.2016.11.022
发表时间: 2016-12-01
期刊: Cell
影响因子: 64.5
作者:
Benci JL;Xu B;Qiu Y;Wu TJ;Dada H;Twyman-Saint Victor C;Cucolo L;Lee DSM;Pauken KE;Huang AC;Gangadhar TC;Amaravadi RK;Schuchter LM;Feldman MD;Ishwaran H;Vonderheide RH;Maity A;Wherry EJ;Minn AJ
通讯作者: Minn AJ
DOI: 10.1501/0000000035
发表时间: 2011-01-01
影响因子: 36.5
作者:
Criminisil, Antonio;Shotton, Jamie;Konukoglu, Ender
通讯作者: Konukoglu, Ender
DOI: 10.1080/01621459.2017.1319839
发表时间: 2018-01-01
影响因子: 3.7
作者:
Wager, Stefan;Athey, Susan
通讯作者: Athey, Susan
DOI: 10.1214/15-aos1321
发表时间: 2015-08-01
影响因子: 4.5
作者:
Scornet, Erwan;Biau, Gerard;Vert, Jean-Philippe
通讯作者: Vert, Jean-Philippe
DOI: --
发表时间: 2004-12
期刊: --
影响因子: --
作者:
P. Xu;F. Jelinek
通讯作者: P. Xu;F. Jelinek