一种基于粗糙-模糊集理论的分类规则挖掘方法  

An Approach of Mining Classification Rules Based on Rough-Fuzzy Set Theories

作　　者：蔡虹[1] 叶水生[1] 张永[1] 

机构地区：[1]南昌航空工业学院计算机系,南昌330063

出　　处：《计算机工程与应用》2006年第2期186-187,214,共3页Computer Engineering and Applications

基　　金：江西省教育厅资助项目(编号:DB200406011);江西省测控中心资助项目(编号:ZX200328002)

摘　　要：提出了一种基于粗糙-模糊集理论的分类规则挖掘方法,以解决信息不完整情况下的推理和决策问题,并给出了该方法的流程图。利用基于粗糙集的特征属性约简算法和基于模糊集的决策规则归纳方法,可以挖掘出样本中隐藏的关联规则,形成决策。最后,将其应用于一个具体的信息系统中,结果令人满意,证明该方法是可行的且是有效的。On purpose of solving the problems of reasoning and decision-making when the given information is not sufficient,an approach of mining classification .rules based on rough-fuzzy set theories is prompted,and its flow chart is also presented.The approach can mine hidden association rules in samples and make decisions using the reduction algorithm based on rough set and the rule induction algorithm based on fuzzy set.And at the end,this approach is applied to a specific information system.The results are satisfactory,which have proved that the approach is practicable and effective.

关 键 词：粗糙集 模糊集 关联规则 属性约简 规则归纳 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论] O159[自动化与计算机技术—计算机科学与技术]