FUAT - A fuzzy clustering analysis tool


BOZKIR A. S. , Sezer E. A.

EXPERT SYSTEMS WITH APPLICATIONS, cilt.40, ss.842-849, 2013 (SCI İndekslerine Giren Dergi) identifier identifier

  • Cilt numarası: 40 Konu: 3
  • Basım Tarihi: 2013
  • Doi Numarası: 10.1016/j.eswa.2012.05.038
  • Dergi Adı: EXPERT SYSTEMS WITH APPLICATIONS
  • Sayfa Sayıları: ss.842-849

Özet

As it is known, fuzzy clustering is a kind of soft clustering method and primarily based on idea of segmenting data by using membership degrees of cases which are computed for each cluster. However, most of the current fuzzy clustering modules packaged in both open source and commercial products have lack of enabling users to explore fuzzy clusters deeply and visually in terms of investigation of different relations among clusters. Furthermore, without a decision maker or an expert, it is hard to decide the number of clusters in fuzzy clustering studies. Therefore, in this study, a desktop software, namely FUAT, is developed to analyze, explore and visualize different aspects of obtained fuzzy clusters which are segmented by fuzzy c-means algorithm. Moreover, to obtain and inform possible natural cluster number, FUAT is equipped with Expectation Maximization algorithm. (C) 2012 Elsevier Ltd. All rights reserved.