M.Sc. Tezi Görüntüleme

      Cluster analysis has an important role in data mining. The objective of clustering analysis is to partition the data set into subsets using their similarities or dissimilarities. In this study, fuzzy c-means clustering algorithm is adapted for directional data. Several methods have been developed for clustering of directional data in the literature. But, approximate results are obtained in those clustering methods. Therefore, these methods lead to undesirable results for very sensitive problems. In the methods of literature, clustering is performed with approximate distances which are calculated by using trigonometric functions. In this study, fuzzy c-means algorithm for directional data (FCM4DD) uses directly angular distance. Thus more consistent results are obtained with FCM4DD clustering algorithm. FCM4DD algorithm is a clustering algorithm which can be used for N dimensional data as well as circular data. In this study, some existing clustering algorithms and FCM4DD algorithm is applied on various numerical examples and obtained results are compared. The results show that FCM4DD algorithm is more consistent, more accuracy and faster.