相斥邻域的覆盖粗糙集实值属性约简  

Numerical attribute reduction of mutex neighborhood covering rough set theory

作　　者：张灵均[1] 徐久成 李双群[1] 李晓艳[1] 

机构地区：[1]河南师范大学计算机与信息技术学院,河南新乡453007

出　　处：《山东大学学报（理学版）》2012年第1期77-82,共6页Journal of Shandong University(Natural Science)

基　　金：国家自然科学基金资助项目(60873104;61040037);河南省科技攻关重点项目(112102210194);河南省教育厅自然科学基金(2008B520019)

摘　　要：粗糙集理论是一种有效的属性约简方法,但不能直接处理实值数据。针对此问题,本文首先介绍了邻域和覆盖的概念,在此基础上构造了覆盖自约简和覆盖间约简(属性约简)算法;然后通过讨论邻域内各样本之间关系,提出了相斥元的定义,相斥元的存在可能导致决策正域计算错误,从而得到不符合数据表实际情况的属性依赖性,因此给出了分解相斥元的方法;最后在四个实值的基因表达数据库上进行了实验,结果表明该属性约简算法是有效的,并相对于现有其他算法具有较高的分类精度。Rough set theory is an effective method of knowledge reduction, but can not directly deal with numerical at- tributes. To address this problem, based on the concepts of neighborhood and cover, the algorithms of covering itself reduction and covering reduction are constructed. And then, by discussing the relationship among neighborhood sam- pies, the mutex in neighborhood samples is defined. The unreasonable positive region caused by mutex, leads to attrib- ute dependency that may not accord with what the real condition dataset reflects, thus the method of decompose mutex neighborhood is proposed. Finally, we report experimental results with four numerical gene expression datasets and compare the results with some other methods. The results prove that the proposed method is effective, and has higher tumor classification accuracy.

关 键 词：相斥元 邻域 覆盖粗糙集 属性约简 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]