登录注册
高级搜索
包含全部检索词
包含精确检索词
包含至少一个检索词
不包含检索词
出现检索词的位置
文章任何位置
位于文章标题
作者
机构
出版物 期刊
会议
发表时间  - 
语言检索范围 不限
不限 英文 中文
文献
期刊
学者
论文查重 首单免费
开题分析
单篇购买
文献互助

Feature selection based on mutual information criteria of max-dependency, max-relevance, and min-redundancy

来自 NCBI

阅读量：

6728

作者：

H Peng，F Long，C Ding

展开

摘要：

Feature selection is an important problem for pattern classification systems. We study how to select good features according to the maximal statistical dependency criterion based on mutual information. Because of the difficulty in directly implementing the maximal dependency condition, we first derive an equivalent form, called minimal-redundancy-maximal-relevance criterion (mRMR), for first-order incremental feature selection. Then, we present a two-stage feature selection algorithm by combining mRMR and other more sophisticated feature selectors (e.g., wrappers). This allows us to select a compact set of superior features at very low cost. We perform extensive experimental comparison of our algorithm and other methods using three different classifiers (naive Bayes, support vector machine, and linear discriminate analysis) and four different data sets (handwritten digits, arrhythmia, NCI cancer cell lines, and lymphoma tissues). The results confirm that mRMR leads to promising improvement on feature selection and classification accuracy.

展开

关键词：

Index Terms- Feature selection mutual information minimal redundancy maximal relevance maximal dependency classification.

DOI：

10.1109/TPAMI.2005.159

被引量：

7477

年份：

2005

收藏 引用 批量引用 分享
全部来源 免费下载 求助全文

通过文献互助平台发起求助，成功后即可免费获取论文全文。

相似文献

参考文献

引证文献

辅助模式

0

引用

文献可以批量引用啦~
欢迎点我试用！

引用