14.2.12.1 ISODATA Clustering

Chapter Contents (Back)
ISODATA Clustering. ISODATA is similar to K-Means, except ISODATA does not assume a given number of clusters.

Selim, S.Z., and Ismail, M.A.,
On the Local Optimality of the Fuzzy ISODATA Clustering Algorithm,
PAMI(8), No. 2, March 1986, pp. 284-288. See also K-Means-Type Algorithms: A Generalized Convergence Theorem and Characterization of Local Optimality. See also Fuzzy C-Means: Optimality of solutions and effective termination of the algorithm. BibRef 8603

Venkateswarlu, N.B., Raju, P.S.V.S.K.,
Fast isodata clustering algorithms,
PR(25), No. 3, March 1992, pp. 335-342.
WWW Version. 0401 BibRef

Kittler, J.V., Pairman, D.,
Optimality of reassignment rules in dynamic clustering,
PR(21), No. 2, 1988, pp. 169-174.
WWW Version. 0309ISODATA. It is shown that contrary to popular belief these iterative clustering algorithms do not guarantee that each stable partition is locally optimal. BibRef

Carman, C.S.[Charles S.], Merickel, M.B.[Michael B.],
Supervising ISODATA with an information theoretic stopping rule,
PR(23), No. 1-2, 1990, pp. 185-197.
WWW Version. 0401 BibRef

Huang, K.Y.[Kai-Yi],
A Synergistic Automatic Clustering Technique (SYNERACT ) for Multispectral Image Analysis,
PhEngRS(68), No. 1, January 2002, pp. 33-40. A new effective synergistic automatic clustering technique serves as a substitute for ISODATA when applied to remote sensing image analysis with a large data set.
WWW Version. 0201 BibRef


Chapter on Pattern Recognition, Clustering, Statistics, Grammars, Learning, Neural Nets, Genetic Algorithms continues in
Nonparametric Clustering .


Last update:Jun 25, 2008 at 13:37:57