Affiliations: Department of Computer Engineering, Ferdowsi University of Mashhad, Mashhad, Iran | Faculty of Mathematics, Ferdowsi University of Mashhad, Mashhad, Iran | Center of Excellence on Soft Computing and Intelligent Information Processing, Ferdowsi University of Mashhad, Mashhad, Iran
Note: [] Corresponding author. Hadi Sadoghi Yazdi, Department of Computer Engineering, Ferdowsi University of Mashhad, Mashhad, Iran. E-mail: h-sadoghi@um.ac.ir
Abstract: This paper presents a new hierarchical tree approach to clustering fuzzy data, namely extensional tree (ET) clustering algorithm. It defines a dendrogram over fuzzy data and using a new distance between fuzzy numbers based on α-cuts. The present work is based on hierarchical clustering algorithm unlike existing methods which improve FCM to support fuzzy data. The Proposed ET clustering algorithm is compared with some of the newly presented methods in the literature. The major advantage of ET, first tree clustering method over fuzzy number, in comparison with other algorithms is its fault tolerance against noisy samples. Some examples prove ability of the proposed ET.
Keywords: Fuzzy data, tree clustering, dissimilarity measure, α-cuts
