基于决策表的模糊粗糙单调依赖算法及其应用  

Fuzzy Rough Monotone Dependence Algorithm Based on Decision Table and Its Application

作　　者：梁瑾[1,2] 罗飞[1] 许玉格[1] 

机构地区：[1]华南理工大学自动化科学与工程学院,广东广州510640 [2]华南师范大学教育信息技术学院,广东广州510631

出　　处：《华南理工大学学报（自然科学版）》2011年第7期7-12,共6页Journal of South China University of Technology(Natural Science Edition)

基　　金：国家自然科学基金资助项目(60774032);广东省自然科学基金博士科研启动项目(9451064101002853);广东省自然科学基金资助项目(10151064101000075);华南理工大学中央高校基本科研业务费专项资金资助项目(2011ZM0126)

摘　　要：为了挖掘输入输出之间量的单调依赖关系,根据模糊粗糙集理论,通过对实践中量依赖关系的数据分析,提出了基于决策表的模糊单调依赖算法.文中通过区间映射定义了模糊单调依赖关系和模糊单调递增依赖隶属关系等概念,讨论了模糊单调递增依赖关系的参数和性质,并证明了相关命题.将该算法应用到工业污水的处理中,结果表明该算法是一种有效的数据挖掘算法.In order to reveal the quantity monotone dependence of the output on the input,a fuzzy monotone dependence algorithm based on decision table is proposed by analyzing the data about the quantity dependence in practice and by applying the fuzzy rough set theory.Then,the concepts such as the fuzzy monotone dependence and the fuzzy monotone increasing dependence membership function are defined by interval mapping,and the parameters and characteristics describing the fuzzy monotone increasing dependence are discussed.Moreover,the correlative proposition is proved.The proposed algorithm is finally applied to the wastewater treatment,with its effectiveness in data mining being verified.

关 键 词：数据挖掘 模糊粗糙集 决策表 模糊单调依赖关系 映射 隶属函数 区间划分 递增依赖 

分 类 号：TP18[自动化与计算机技术—控制理论与控制工程]