粗糙集理论的连续属性离散化的一种方法

A Method to Discretization for Successive Attribute of Rough Set Theory

文中提出了基于属性支持度的连续属性的离散化的一种方法 ,在Na veScaler算法基础上利用了决策属性支持度作为准则 ,进一步减少了用Na veScaler算法处理后的离散化后的断点数 ,试验表明了该方法使得离散化后的规则集合更加简洁 ,更便于属性约简。

The paper presents a method of discretization of successive attribute based on decision attribute support degree. Based on Nave Scaler algorithm,we take decision attribute support degree as a regulation,which can decrease the segmentation dot number after it is dealt with by Nave Scaler algorithm. The experiment demonstrates that this method is more concise for the rule set after discretization and more convenient for attribute reduction.