为了实现模糊信息系统属性约简问题,探讨一种基于正域类的模糊属性约简方法。利用模糊二元相似关系建立模糊相似类的方法,进而定义决策属性关于条件属性的模糊正域类,得到正域类属性的重要度。结果表明:一方面,实现模糊信息系统属性约简;另一方面,通过取不同的相似精度,得到不同属性约简集。
In order to achieve attribute reduction in fuzzy information system, a fuzzy attribute reduction method is discussed based on the fuzzy positive domain. Fuzzy similarity class is established using fuzzy binary similarity relation methods. Thereby a decision attribute on condition attribute fuzzy regular domain classes is established and positive domain class attribute importance extent is obtained. The results show that: on the one hand, fuzzy information system attribute reduction is achieved; On the other hand, different similarity precision is taken and a different set of attribute reduction is gotten.