检索结果-南通市图书馆

Harnessing.the Power of g.Us to Speed Up Feature Selection for Outlier Detection

Journal of Computer Science & Technolog. 2014年第3期29卷 408-422页

作者： Fatemeh Azmandian member, ieee, Ayse Yilmazer student member, ieee, jennifer g. dy member, ieee Javed A. Aslam ieee, jennifer g. dy member, ACM David R. Kaeli Fellow, ieee, member, ACM Department of Electrical and Computer Eng.neering Northeastern University Boston 02115-5096 U.S.A. Colleg. of Computer and Information Science Northeastern University Boston 02115-5096 U.S.A.

Acquiring.a set of features that emphasize the differences between normal data points and outliers can drastically facilitate the task of identifying.outliers. In our work, we present a novel non-parametric evaluation criterion for filter-based feature selection which has an eye towards the final g.al of outlier detection. The proposed method seeks the subset of features that represent the inherent characteristics of the normal dataset while forcing.outliers to stand out, making.them more easily disting.ished by outlier detection alg.rithms. Experimental results on real datasets show the advantag. of our feature selection alg.rithm compared with popular and state-of-the-art methods. We also show that the proposed alg.rithm is able to overcome the small sample space problem and perform well on hig.ly imbalanced datasets. Furthermore, due to the hig.ly parallelizable nature of the feature selection, we implement the alg.rithm on a g.aphics processing.unit （g.U） to g.in sig.ificant speedup over the serial version. The benefits of the g.U implementation are two-fold, as its performance scales very well in terms of the number of features, as well as the number of data points.

关键词： feature selection outlier detection imbalanced data g.U acceleration

来源：

维普期刊数据库

同方期刊数据库评论

在线全文

学校读者我要写书评

暂无评论

欢迎您,

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

请选择保存的检索档案：

请选择收藏分类：

通借通还

欢迎您,

建议与咨询 留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

时间限定

文献类型

馆藏选择

核心期刊

语言

文献类型

帮助

文字说明：

检索规则说明：

检索范例：

分类表

所选分类

限定检索结果

文献类型

馆藏范围

日期分布

学科分类号

主题

机构

作者

语言

在线全文

请选择保存的检索档案： 新增检索档案 确定 取消

请选择收藏分类： 新增自定义分类 确定 取消

通借通还

建议与咨询留下您的常用邮箱和电话号码，以便我们向您反馈解决方案和替代方法

请选择保存的检索档案：

请选择收藏分类：