基于卡方距离的直觉模糊相似度聚类算法

A Study of Intuitionistic Fuzzy Similarity Clustering Algorithm Based on Chi-Square Distance

卡方距离作为一类重要的信息测度,在统计模式识别和图像处理领域有成功的应用。经典卡方距离是为度量2个精确数向量而提出,但针对直觉模糊数信息,还未有文献进行讨论。针对直觉模糊集情形,首先提出了一类基于卡方距离的直觉模糊相似度测度,其次在新的相似度基础上提出了一类新的直觉模糊聚类方法,进而得到相似矩阵对备选对象集进行聚类,最后通过算例验证了所提出的聚类方法的有效性和可行性。

Chi-square distance,as an important information measure,has been successfully applied in statistical pattern recognition and image processing.The classical chi-square distance is proposed to measure two exact number vectors,but for intuitionistic fuzzy number information,there is no literature to discuss.Firstly,for the case of intuitionistic fuzzy sets,a class of intuitionistic fuzzy similarity measure based on Chi-square distance is proposed.Secondly,based on the new similarity,a new kind of intuitionistic fuzzy clustering method is proposed.Then the similarity matrix is obtained to cluster the objects.Finally,an example is given to verify the effectiveness and feasibility of the proposed clustering method.