基于粗集理论的分类关联规则挖掘研究

认为数据量的巨大和高维、用户交互与先验知识的利用等等是知识发现领域面临的问题和难点 .粗糙集理论是一种具有模糊边界的集合理论 ,它作为研究知识发现的新型工具 ,能严格地处理不精确数据的分类问题 ,被广泛应用于不相容决策表中的规则提取过程中 .针对粗糙集理论中属性约减和属性值约减这两个重要问题进行了研究 ,并介绍了数据集中挖掘分类规则的基本原理 ,同时利用 RS理论中核和核值的概念 ,提出了一个在数据集中发现没有冗余属性的最小归纳依赖关系并简化决策系统的数据挖掘算法 ,并应用一简单的例子说明如何在数据库中发现分类规则 ,实验结果表明此算法可以大大提高系统潜在知识的清晰度

Research on Data Mining Classification Rule Based on Rough Set Theory

In order to improve the precision and interpretation ability of data mining, the authors study two important issues of attribute reduction and property value reduction in rough set theory. And based on the study, an algorithm of data mining is put forward, and this algorithm can find the conditional attributes related to decision attributes in data set. Through an example, it is illustrated how to discover classification rule in database, and the result of the example shows that the algorithm can greatly improve the resolution of knowledge.