| 一种基于开方检验的特征选择方法 |
| |
| 引用本文： | 黄 源,李 茂,吕建成. 一种基于开方检验的特征选择方法[J]. 计算机科学, 2015, 42(5): 54-56, 77 |
| |
| 作者姓名： | 黄 源 李 茂 吕建成 |
| |
| 作者单位： | 四川大学计算机科学学院 成都610065 |
| |
| 基金项目： | 本文受教育部博士点基金资助 |
| |
| 摘 要： | 开方检验是目前文本分类中一种常用的特征选择方法.该方法仅关注词语和类别间的关系,而没有考虑词与词之间的关联,因此选择出的特征集具有较大的冗余度.定义了词语的“剩余互信息”概念,提出了对开方检验的选择结果进行优化的方法.使用该方法可以得到既有很强表征性又有很高独立性的特征集.实验表明,该方法表现良好.
|
| 关 键 词： | 文本分类 特征选择 开方检验 互信息 |
New Feature Selection Method Based on CHI |
| |
| HUANG Yuan,LI Mao and LV Jian-cheng. New Feature Selection Method Based on CHI[J]. Computer Science, 2015, 42(5): 54-56, 77 |
| |
| Authors: | HUANG Yuan LI Mao LV Jian-cheng |
| |
| Affiliation: | College of Computer Science,Sichuan University,Chengdu 610065,China,College of Computer Science,Sichuan University,Chengdu 610065,China and College of Computer Science,Sichuan University,Chengdu 610065,China |
| |
| Abstract: | CHI is a widely used feature selection method in text classification.This method only focuses on the relevance between features and classifications but ignores the relevance between feature and feature,resulting in a high redundancy.This paper proposed a concept about residual mutual information,and then CHI and residual mutual information were combined together to optimized the selective results.The experimental results indicate that the method is effective. |
| |
| Keywords: | Text categorization Feature selection CHI Mutual information |
| 本文献已被 万方数据 等数据库收录！ |
|
点击此处可从《计算机科学》下载全文 |
|
