DocumentCode
3180273
Title
A non-linear K-means algorithm and its application to unsupervised clustering
Author
Yu, Yong ; Trouvé, Alain
Author_Institution
Departement TSI, Ecole Nat. Superieure des Telecommun., Paris, France
Volume
2
fYear
2002
fDate
26-30 Aug. 2002
Firstpage
1146
Abstract
A new partition criterion for pairwise clustering is proposed naturally in the probabilistic analysis framework. Its connection to the normal K-means algorithm is explained in two different views which also builds its relationship with the kernel approach introduced by Vapnik. Both synthetic examples and the challenging task of planar shape analysis have been given to show its efficiency in unsupervised pairwise clustering application.
Keywords
database theory; pattern clustering; probability; tree data structures; database; hierarchical clustering tree; kernel approach; nonlinear K-means algorithm; partition criterion; planar shape analysis; probabilistic analysis framework; unsupervised pairwise clustering; Clustering algorithms; Content based retrieval; Image converters; Image databases; Image retrieval; Information retrieval; Kernel; Partitioning algorithms; Pattern recognition; Shape;
fLanguage
English
Publisher
ieee
Conference_Titel
Signal Processing, 2002 6th International Conference on
Print_ISBN
0-7803-7488-6
Type
conf
DOI
10.1109/ICOSP.2002.1179992
Filename
1179992
Link To Document