一种基于粗糙集分类规则发现的方法

Method of discovering classification rule based on rough set theory

介绍了粗糙集理论实现数据分类和规则推理的基本原理,并利用粗糙集理论中核及决策类覆盖的概念,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,并简化带有不相容规则的决策系统的多层次数据挖掘算法,应用一实际的例子说明如何在数据库中发现分类规则。

Introduces the basic principle for rule generation and mining classification rules in data sets and describes the concept of core and decision class covering in rough set theory.Then,by analyzing induction dependency relation among attributes,we give a new method of discovering minimum induction dependency relation in data sets.Which is used in mine multi-layer classification rules with non-consistent decision rules in a decision system.In the end,using the method on a realistic database shows it is effective.