Hierarchical Clustering of Imprecise Data

Clustering is an important data mining technique that possesses immense applications in many research areas. However, majority of the clustering algorithms available are focused towards data sets that contain precise values. But there are situations in which one can come across data sets whose objects take values that are imprecise or fuzzy in nature. To model imprecise data, various mathematical theories have been propounded in the literature. One such theory that is recently developed is Credibility Theory. In this paper, a new hierarchical clustering algorithm for objects that take imprecise values is developed by using some of the concepts available in Credibility Theory. A numerical illustration of the proposed algorithm is also provided.