一种基于粗糙集的单一属性分类的约简方法

An Reduction Approach for Single Attribute Class Based on Rough Sets Theory

粗糙集的应用中 ,对象集通常具有数量大、属性多、单一属性分类大的特点 ,这是在已有知识基础上提出规则时所遇到的 3个主要问题 .针对其中单一属性分类的约简问题提出了一种新的方法 ,它采用了先合并后分解再综合的思想简化问题的求解 。

During the application of the Rough sets, the object sets usually have a character of large quantity, attributes and single attribute class. These problems appear when the regulation is proposed based on the past knowledge. With the view of single attribute class, this paper proposes a new method which is combining class firstly, then discomposing, last synthesizing. This way can simplify the solution and mine the regulation in the information quickly and sufficiently.