摘要
基于属性约简的方法,放弃以往复杂的规则匹配算法,提出将约简后的多种属性组进行析取,筛选特征项,并构造分类器.实验结果表明,此算法不仅简单,还能降低维数和提高分类结果.
Based on attributes reduction and discarding old complicated algorithms of matching rules, the method of combining several attributes set after attribute reduction and filtering the attributes, then constructing classifiers is presented. The results show that this way is not only easy, but also reduces the dimension and advances the results of categorization.
出处
《郑州大学学报（理学版）》
CAS
2007年第2期100-103,共4页
Journal of Zhengzhou University:Natural Science Edition
基金
国家自然科学基金资助项目
编号60373095
60673039
国家高科技863计划项目
编号2006AA01Z151
关键词
文本分类
向量空间模型
粗糙集
属性约简
text categorization
vector space model
rough set
reduction of attribute
