作　　者： ; ;

机构地区： 华中科技大学计算机科学与技术学院

出　　处： 《华中师范大学学报（自然科学版）》 2005年第1期20-23,共4页

摘　　要： 模糊似然函数是刻画两个模糊集之间相似程度的重要工具,众多学者给出了多种形式的计算方法,但由于实际问题的复杂性,有些方法在某些情况下会产生违反直觉的结论.首先就闫德勤在2001年建议的方法进行讨论,然后证明了一个引理,提出一种新的模糊似然函数,其本质是由模糊集合间lp距离导出的贴近度.与相关公式比较,发现它们能导出相同的模糊熵,最后给出一个它在模式识别中的应用实例. Fuzzy likelihood function is one of the important tools to describe the degree of similarity between fuzzy sets. Lots of researchers have developed many methods. However, some of them can result against our feeling in some special cases because of the complexity of practical problems. At first, the method proposed by Yan Deqing in 2001 is discussed and its drawbacks are pointed out . A lemma is proved and then a new fuzzy likelihood function is presented. It is just a nearness degree derived from the l\+p distance between fuzzy sets. Compared with related formulations, it is found that they can induce the same fuzzy entropy. Finally, an application example of it in pattern recognition is showed.

关 键 词： 模糊似然函数 距离 模糊熵 模式识别

领　　域： [自动化与计算机技术] [自动化与计算机技术] [理学] [理学]