Presentation 1995/7/27
Competitive Models for Unsupervised Clustering
Ryuta Itoh, Takehiko Shida, Toshiki Kindo,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) In this paper, three competitive models, which estimate the prototype of cluster with unsupervised clustering, are proposed. The first one is LVQ+MM model, which is composed of the learning vector quantization (LVQ) and the MAX-MIN distance method. The second one is the learning biased vector quantization (LBVQ) model with the extened distance measure which is added a biased term to that of LVQ. The third one is the learning Bayesian quantization (LBQ) model whose distance measure is given by discriminant function of the Bayes decision. The numerical experiment with pseudo-random numbers for two-dimensional input pattern whose probability density function is normal distribution shows the LBQ model is the best model to estimate the prototype of cluster and the incidence probability of cluster in these models. The LBQ model gives the Bayes decision boundary surface of each real cluster approximately.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) competition / clustering / prototype / Bayes decision
Paper #
Date of Issue

Conference Information
Committee NC
Conference Date 1995/7/27(1days)
Place (in Japanese) (See Japanese page)
Place (in English)
Topics (in Japanese) (See Japanese page)
Topics (in English)
Chair
Vice Chair
Secretary
Assistant

Paper Information
Registration To Neurocomputing (NC)
Language JPN
Title (in Japanese) (See Japanese page)
Sub Title (in Japanese) (See Japanese page)
Title (in English) Competitive Models for Unsupervised Clustering
Sub Title (in English)
Keyword(1) competition
Keyword(2) clustering
Keyword(3) prototype
Keyword(4) Bayes decision
1st Author's Name Ryuta Itoh
1st Author's Affiliation Matsushita Research Institute Tokyo, Inc.()
2nd Author's Name Takehiko Shida
2nd Author's Affiliation Matsushita Research Institute Tokyo, Inc.
3rd Author's Name Toshiki Kindo
3rd Author's Affiliation Matsushita Research Institute Tokyo, Inc.
Date 1995/7/27
Paper #
Volume (vol) vol.95
Number (no) 189
Page pp.pp.-
#Pages 8
Date of Issue