粗糙集-决策树在数据挖掘中的应用  

Applications of Data Mining Approach Based on Rough Sets and Decision Tree

作　　者：吴成东[1] 许可[1] 张海波[1] 刘建顺[1] 李旸[1] 

机构地区：[1]沈阳建筑大学信息与控制工程学院,辽宁沈阳110168

出　　处：《沈阳建筑大学学报（自然科学版）》2005年第4期386-389,共4页Journal of Shenyang Jianzhu University：Natural Science

基　　金：科技部国际合作重点项目(2003DF020009)

摘　　要：目的将数据挖掘技术应用在胶合板缺陷检测数据中,提取出有效的、正确的规则信息.方法通过分析比较粗糙集软计算方法和决策树方法的特点,利用两种方法具有的优势互补性,将其进行有机集合,构造数据挖掘模型.结果从胶合板缺陷检测数据中挖掘出对用户有价值的决策规则,并将其用“IF-THEN”语句表达出来,以便指导以后的决策过程.结论基于粗糙集和决策树结合的数据挖掘方法提高了获取规则的快速性,降低了计算的复杂度,增强规则的可解释性,取得了良好的研究结果.In order to obtain effective, right information from the data of the defect inspection of wood veneer, the method is studied based on data mining. The characteristics of rough sets soft computing and decision tree are analyzed and Compared. The paper combines rough sets with decision tree through their mutual complement in superiority. The model of data mining is built. As a result, the method discovers valuable rule from the data of the defect inspection of wood veneer, which is expressed by the pattern “if-then” for users, so as to instruct the later decision process. Data mining approach based on rough sets and decision tree improves the speed of discovering rules, reduces the complexity of computing, and enhances the possibility of being explained. Therefore, a satisfactory research result has been made.

关 键 词：数据挖掘 粗糙集 属性约简 决策树 

分 类 号：TP391[自动化与计算机技术—计算机应用技术]