基于模糊贴近度的属性约简

Attribute Reduction based on Fuzzy Closeness Degree

属性约简是信息系统中知识发现的重要内容.基于模糊决策信息系统探讨模糊环境下的属性约简问题.首先根据模糊贴近度定义模糊决策信息系统中的决策协调集与决策约简集,由此引入一种关于贴近度的属性辨识集,然后结合属性的特征讨论了模糊决策信息系统的属性约简理论,提出一种基于模糊贴近度的属性约简方法,并举例验证和说明此方法的运用。

Attribute reduction is an important content of knowledge discovery in information systems. This paper first in- troduces the notion and properties of fuzzy nearness degree. Then, According to the fuzzy nearness the decision coordi- nation set and the decision reduction set for the fuzzy decision information system are defined. Furthermore, the attribute discernibility set is established by the fuzzy nearness degree, and meanwhile, the attribute reduction of fuzzy decision in- formation system is discussed. As a result, a new attempt of attribute reduction theories and methods based on fuzzy nearness degree is proposed in fuzzy information systems, and an example is given to illustrate the application of this method.