A New Dimensionality Reduction Method with Correlation Analysis and Universum Learning

A New Dimensionality Reduction Method with Correlation Analysis and Universum Learning
复制标题

一种新的关联分析和宇宙学习的降维方法

DOI:
10.1134/s1054661818020189
复制
发表时间:
2018-04
影响因子:
1
通讯作者:
Wang Liping
Wang Liping
中科院分区:
--
文献类型:
--
作者:
Chen Xiaohong;Wang Liping

文献摘要

参考文献

相似文献

先验性约简(DR)是机器学习中一个重要且必不可少的预处理步骤,可能使用区分信息、近邻信息或相关信息,从而产生不同的DR方法。在这项工作中,我们设计了新的DR方法,采用另一种形式的信息,即,Universum数据上的最大矛盾,这些数据与手头的任务属于同一个域,但不属于任何类别的训练数据。已经发现,在Univesum数据的帮助下,分类和聚类算法实现了有利的改进,这种学习方法被称为Univesum学习。 这两类问题都可以用广义特征值问题来表示,并通过特征值计算来求解。在合成数据集和真实数据集上的实验表明,所提出的DR方法可以获得更好的性能。
Dimensionality reduction (DR) is an important and essential preprocessing step in machine learning, possibly using discriminative information, neighbor information or correlation information and resulting in different DR methods. In this work, we design novel DR methods that employ another form of information, i.e., the maximal contradiction on Universum data which belong to the same domain as the task at hand, but do not belong to any class of the training data. It has been found that classification and clustering algorithms achieve favorable improvements with the help of Universum data and such learning methods are referred as to Univesum learning. Both of them can be expressed by generalized eigenvalue problem and solved by eigenvalue computation. The experiments on both synthetic and real-world datasets are presented to show that the proposed DR methods can obtain better performance.
DOI: 10.1198/jasa.2003.s308
发表时间: 2003-12
影响因子: 3.7
作者:
M. Timmerman
通讯作者: M. Timmerman
DOI: 10.1002/wics.101
发表时间: 2010-07-01
影响因子: 3.2
作者:
Abdi, Herve;Williams, Lynne J.
通讯作者: Williams, Lynne J.
DOI: --
发表时间: 2007-12
期刊: --
影响因子: --
作者:
Fabian H Sinz;O. Chapelle;Alekh Agarwal;B. Scholkopf
通讯作者: Fabian H Sinz;O. Chapelle;Alekh Agarwal;B. Scholkopf
DOI: 10.1007/978-1-4612-4380-9_14
发表时间: 1936-12
期刊: Biometrika
影响因子: 2.7
作者:
H. Hotelling
通讯作者: H. Hotelling
DOI: --
发表时间: 1996
期刊: --
影响因子: --
作者:
C. Merz
通讯作者: C. Merz