摘要
现有规则提取方法大多数只能在相容决策系统中提取规则,并且提取出的规则冗余度高、用户不易理解。针对该问题,提出一种基于对象集覆盖的规则提取方法,利用粗糙集理论将对象集划分为相应的等价类,根据属性特征值的一致性程度和相似程度产生有效性规则,通过等价类划分和对象集覆盖解决不相容决策系统的规则提取问题。算例分析结果表明,该方法提取出的规则简单可靠,具有较好的鲁棒性。
The current rules extraction methods mostly only extract rules in consistent decision system,and the extracted rules redundant degree is high,it is not easy to understand for users.Aiming at this problem,this paper puts forward a rule extraction method based on cover object set.It divides object set into corresponding equivalence class by using rough set theory,produces effectiveness rules according to the consistency of the similarity degree,and solves inconsistent decision system rules extraction through the equivalence partitioning and object set cover.Numerical example analysis result proves the validity and reliability of the method,and it has good robustness.
出处
《计算机工程》
CAS
CSCD
2013年第2期46-49,共4页
Computer Engineering
基金
安徽省自然科学基金资助项目(090412054)
安徽省科技攻关计划重大科技专项基金资助项目(08010201002)
安徽高等学校省级自然科学基金资助项目"不完备信息系统中基于粗糙集的知识获取研究"(KJ2011Z020)
关键词
数据挖掘
粗糙集
不相容决策系统
规则提取
特征集
决策表
data mining
rough set
inconsistent decision system
rule extraction
feature set
decision table
