Relational visual cluster validity

摘要: The assessment of cluster validity plays a very important role in analysis. Most commonly used methods are based on statistical hypothesis testing or finding the best clustering scheme by computing number different indices. A visual have been produced to display directly clusters mapping data into two- three-dimensional space. However, these may lose too much information correctly estimate results algorithms. Although (VCV) method Hathaway and Bezdek can successfully solve this problem, it only be applied for object data, i.e. feature measurements. There few that analyze where similarity dissimilarity relation exists – relational data. To tackle paper presents (RVCV) assess This is done combining non-Euclidean fuzzy c-means (NERFCM) algorithm with modification VCV produce representation validity. RVCV complete incomplete adds theory. Numeric examples using synthetic real presented