中介粗集及其在数据挖掘中的应用

粗集理论已成功应用于数据挖掘，如基于粗集方法的属性发现和决策规则的约简等。中介集合论是以中介逻辑为基础，也可以用于数据挖掘。文中将两者相结合，提出了中介粗集的概念。中介粗集是普通粗集的推广，为不完全信息系统的数据挖掘提供了新的工具，不仅可用于一般的信息系统，还适用于数据取空值的信息系统（允许结论取空值）。利用中介粗集进行数据挖掘不但可以得出粗集理论中的约简，还能得到更为简洁的中介约简。文中还给出了

Medium Rough Set and Its Application to Data Mining

Rough set (RS) theory introduced by Z.Pawlak is a mathematical approach to data mining and has been successfully applied to data mining in medicine and industry. Medium set (MS) thoery based on medium logic can also be used for data mining.This paper introduces medium rough set (MRS) which is a new tool for mining data from incomplete information system.MRS is an extension of RS,it allows null value among the data(decision attribute value). This paper also gives some properties of MRS and data reductions based on MRS. Using MRS,we can get not only the traditional reduction but also terser medium reduction from incomplete infomation system,so MRS can help decision maker to find the key information in less time. We are studing the generalized MRS which is directly based on MS and can well deal with all the null value in the information system.