数据挖掘中一种基于粗糙集理论的属性值离散映射方法

Research on the Method of Attribute Value Discrete Mapping in Data Mining

不确定性知识是数据挖掘的一个重要内容 ,而属性值的离散映射是在数据库中发现不确定性知识的一个重要环节。本文首先提出一个不确定性信息系统的粗糙集模型 ,在此基础上进一步提出了划分属性值离散区间的算法。该算法同时适用于字符型属性的数据和数值型属性的数据 ,从而可以广泛地应用于不确定性知识的挖掘。

Uncertainty knowledge is an important content of data mining,and attribute value discrete mapping is a key step in mining uncertainty knowledge.In this paper,a rough set model of uncertainty information system is proposed,then a method of mapping attribute's value into discrete value with a fine value partition granularity is developed.This method is suitable for continuous numerical value as well as character value,so it can be widely applied to mining uncertainty knowledge.