摘要
针对大量连续属性值的数据挖掘,提出了一种基于模糊工具箱和ROSETTA软件的粗糙集数据挖掘方法.在粗糙集理论的基础上,应用模糊工具箱中的模糊聚类方法离散分类连续属性值,并将其转化为粗糙集易于处理的知识表格.应用粗糙集数据挖掘软件ROSETTA对这些知识表格进行知识约简处理.通过约简知识属性和属性值,得到连续属性值的核心知识规则,并以实测数据为例,说明了该方法的实现过程和有效性.
Method based on fuzzy toolbox and ROSETTA with rough set to data mining of mass continuous attribute value is proposed. Fuzzy clustering of fuzzy toolbox in existence disperses continuous attribute value and transform it to knowledge sheets which are easy for rough set dealing with. These knowledge sheets are predigested by rough set application software-ROSETTA for data mining. After attributes value and knowledge attributes are predigested, kernel knowledge rulers of the continuous attribute values are obtained. The implement steps and validity of method is explained with the actual data.
出处
《微计算机信息》
北大核心
2007年第18期174-175,178,共3页
Control & Automation
基金
国家自然科学基金委员会创新研究群体科学基金资助(50421703)
