摘要： 模糊K-Modes聚类算法是对具有分类属性的数据进行聚类的一种有效的算法。为了评价聚类结果，以具有明确分类结构的数据作为输入数据，将模糊K-Modes聚类结果与原始数据的分类结构进行对比，分析了确定它们之间对应关系的方法，在期望聚类结果应该具有的特点的基础上，对现有的精确度定义和计算方法进行修正，在划分相似度的基础上，重新定义模糊K-Modes聚类精确度。

关键词: 模糊K-Modes聚类, 精确度, 分类属性, 相似度

Abstract: The fuzzy K-Modes clustering algorithm is an effective method for clustering the data with categorical attributes. Considering the clustering result, this paper uses the classified data, Soybean disease data set, as the input samples, contrasts the clustering result with the classification of the data, analyses the methods to obtain their corresponding relation, modifies the present definition and computing method of the clustering accuracy. Based on the similarity of the different partitions, a new definition of the fuzzy K-Modes clustering accuracy is presented.

中图分类号: 

• TP181