-
- Endo Yasunori
- University of Tsukuba
-
- Tokuyama Haruki
- University of Tsukuba
-
- Miyamoto Sadaaki
- University of Tsukuba
Bibliographic Information
- Other Title
-
- 球面k-mean++法について
- 球面k-means++法について
- キュウメン k-means++ホウ ニ ツイテ
Search this article
Abstract
k-means clustering (KM) algorithm, also called hard c-means clustering (HCM) algorithm, is a very powerful clustering algorithm, but it has a serious problem of strong initial value dependence. To decrease the dependence, Arthur and Vassilvitskii proposed an algorithm of k-means++ clustering (KM++) algorithm on 2007. By the way, there are many case that each object is allocated on an unit sphere, e.g. text clustering. Honik, Kober, and Buchta proposed spherical k-means clustering (SKM) algorithm to classify such objects on 2012. However, the algorithm also has the same problem of initial value dependence as KM. Therefore, this report discuss the following points: (1) the dissimilarity of SKM is extended to satisfy the triangle inequality, and (2) spherical k-means++ clustering (SKM++) algorithm which works well for the problem is proposed. The report shows that the effectiveness of SKM+ is theoretically guaranteed.
Journal
-
- Proceedings of the Fuzzy System Symposium
-
Proceedings of the Fuzzy System Symposium 31 (0), 81-84, 2015
Japan Society for Fuzzy Theory and Intelligent Informatics
- Tweet
Details 詳細情報について
-
- CRID
- 1390282680649847552
-
- NII Article ID
- 130005488336
-
- NII Book ID
- AA12165648
-
- ISSN
- 18820212
-
- NDL BIB ID
- 026775886
-
- Text Lang
- ja
-
- Data Source
-
- JaLC
- NDL
- CiNii Articles
-
- Abstract License Flag
- Disallowed