摘要
属性值约简是粗集理论的核心内容之一。将属性值约简和数据挖掘相结合,给出支持度、置信度、覆盖度的定义。在此基础上,提出了一种新的属性值约简算法—基于规则综合质量的属性值约简算法.通过算法复杂度分析说明,该算法在一定程度上解决了属性值约简的NP难问题。实例仿真表明该算法在解决一些相关实际问题方面是可行的,具有一定的实用价值。
The attributive value reduction is one of the highlight of rough set theory. This paper associates attributive value reduction with data mining and proposed three concepts: support, confidence and coverage. Based on this, a new attributive value reduction algorithm based on overall quality is presented. The algorithm complexity analysis shows that, to a certain extent, the algorithm could resolve the NP hard problems of attributive value reduction. Simulation example shows that the algorithm in solving some related practical issues it is feasible and has a certain practical value.
出处
《计算机与数字工程》
2009年第2期1-3,共3页
Computer & Digital Engineering
基金
四川省科技计划项目(编号:2008GZ0003)资助
四川省科技厅科技攻关项目(编号:07GG006-014)资助
中国科学院人才培养计划项目("西部之光")资助
关键词
属性值约简
粗集
算法复杂度
NP难问题
attributive value reduction, rough set, algorithm complexity, NP hard problem