VDM算法在数据采掘中的应用

VDM Algorithm in the Application of Data Mining

介绍了在数据库知识发现 (KDD)中将连续属性离散化的一些方法 ,并提出使用值差分度量离散化的算法 .值差分度量算法原本是用于计算离散属性值间的距离 ,但实际上将这种方法反过来用于连续属性的离散化也可以有相当好的效果 .将其与传统的使用统计量 χ2 的离散化算法作了比较 .

Some methods for dividing continuous attributes in KDD (knowledge discovery in database) and a method based on VDM (value difference metric) are presented. VDM is used in calculating distance of continuous attributes and in continuous attribute discretion and has a good effect. The comparison between VDM and χ 2 statistical method is also presented.