Simultaneous clustering and dimensionality reduction using variational bayesian mixture model

Kazuho Watanabe, Shotaro Akaho, Shinichiro Omachi, Masato Okada

Research output: Chapter in Book/Report/Conference proceedingConference contribution

2 Citations (Scopus)

Abstract

Exponential principal component analysis (e-PCA) provides a framework for appropriately dealing with various data types such as binary and integer for which the Gaussian assumption on the data distribution is inappropriate. In this paper, we develop a simultaneous dimensionality reduction and clustering technique based on a latent variable model for the e-PCA. Assuming the discrete distribution on the latent variable leads to mixture models with constraint on their parameters. We derive a learning algorithm for those mixture models based on the variational Bayes method. Although intractable integration is required to implement the algorithm, an approximation technique using Laplace's method allows us to carry out clustering on an arbitrary subspace. Numerical experiments on handwritten digits data demonstrate its effectiveness for extracting the structures of data as a visualization technique and its high generalization ability as a density estimation model.

Original languageEnglish
Title of host publicationClassification as a Tool for Research - Proceedings of the 11th IFCS Biennial Conference and 33rd Annual Conference of the Gesellschaft fur Klassifikation e.V., GfKl 2009
Pages81-89
Number of pages9
DOIs
Publication statusPublished - 2010 Dec 1
Event11th Biennial Conference of the International Federation of Classification Societies, IFCS 2009 and with the 33rd Annual Conf of the German Classification Society (Gesellschaft fur Klassifikation) on Classification as a Tool fo Research, GfKl 2009 - Dresden, Germany
Duration: 2009 Mar 132009 Mar 18

Publication series

NameStudies in Classification, Data Analysis, and Knowledge Organization
ISSN (Print)1431-8814

Other

Other11th Biennial Conference of the International Federation of Classification Societies, IFCS 2009 and with the 33rd Annual Conf of the German Classification Society (Gesellschaft fur Klassifikation) on Classification as a Tool fo Research, GfKl 2009
CountryGermany
CityDresden
Period09/3/1309/3/18

ASJC Scopus subject areas

  • Computer Science Applications
  • Information Systems
  • Information Systems and Management
  • Analysis

Fingerprint Dive into the research topics of 'Simultaneous clustering and dimensionality reduction using variational bayesian mixture model'. Together they form a unique fingerprint.

Cite this