摘要： 为解决Vague值(集)的相似度量问题，提出一种公理化意义下全面综合的度量方法。通过对现有Vague集相似度量的功能进行分析对比，获得在相似度量公理化下的扩展形式，并结合模糊熵理论，综合Vague值(集)相对熵的公理化思想，研究Vague值(集)相似度量与相对熵的关系，给出带参数的计算方法。分析结果证明，将该方法应用于国画类别的区分识别中，能获得较好的分类结果。

关键词: Vague值(集), 相似度量, 模糊熵, 相对熵, 模拟, 类别区分

Abstract: As one of the important issues in artificial intelligence, uncertainty measuring methods are put forward and widely used in applications. Considering this, this paper introduces several existing definitions about similarity measure on Vague values(sets), along with the analyzing of their properties. Then it presents an effective matching definition of similarity measure between Vague values(sets). Further, it proposes a formula to compute the entropy of Vague values(sets) together with an axiomatizing idea on fuzzy entropy. All the measures between Vague values(sets) are validated with experimental data. The proposed method is improved and the simulation results show the actual effect on category recognition of Chinese drawing.

Key words: Vague values(sets), similarity measure, fuzzy entropy, relative entropy, simulation, category recognition

中图分类号: 

• TP391.41