A Bisecting K-Medoids clustering Algorithm Based on Cloud Model. Issue 11 (2018)

• Record Type: Journal Article
• Authors: Sun, D.; Fei, H.; Li, Q.
• Is Part Of: IFAC-PapersOnLine. Volume 51:Issue 11(2018)
• Page Start: 308
• Page End: 315
• Publication Date: 2018
• Subjects: cluster algorithm -- Connected TV -- user behavior -- bisecting K-Medoids -- cloud similarity; Automatic control -- Periodicals; 629.805
• DOI: 10.1016/j.ifacol.2018.08.301
• Languages: English
• ISSNs: 2405-8963

Abstract

Abstract: In this study, an advanced K-Medoids clustering algorithm has been developed by using an auto-stopped bisecting scheme and replacing the traditional Euclidean distance by cloud similarity. The proposed algorithm is dedicated to cluster the users of Connected TV according to their behavior involved with connected TVs. According to the experimental results, the solutions obtained by the proposed algorithm are quite encouraging and the clustering results are much more stable than Euclidean-distance-based clustering method. The result shows that the proposed method has achieved good results in terms of cluster accuracy and time.