Presentation 2007-10-18
Variational Bayesian Clustering Method Using Mixture of Exponential Family Distributions
Kazuho WATANABE, Shotaro AKAHO, Masato OKADA,
PDF Download Page PDF download Page Link
Abstract(in Japanese) (See Japanese page)
Abstract(in English) Various data types including binary or integer are encountered in some real-world data modeling problems. In such cases, the Gaussian assumption on the data distribution may be inappropriate. Mixtures of exponential family distributions enable to carry out clustering data of different types. Although a learning algorithm based on variational Bayes was derived in a general framework for this model, some detailed calculations are necessary to apply specific practically important mixtures to this framework. This report gives the variational Bayesian algorithms for mixtures of specific exponential family distributions such as binomial, Poisson and exponential. Approximation scheme using Laplace's method is also presented.
Keyword(in Japanese) (See Japanese page)
Keyword(in English) Mixture Model / Exponential Family / Variational Bayes / Laplace Approximation
Paper # NC2007-35
Date of Issue

Conference Information
Committee NC
Conference Date 2007/10/11(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) Variational Bayesian Clustering Method Using Mixture of Exponential Family Distributions
Sub Title (in English)
Keyword(1) Mixture Model
Keyword(2) Exponential Family
Keyword(3) Variational Bayes
Keyword(4) Laplace Approximation
1st Author's Name Kazuho WATANABE
1st Author's Affiliation Department of Complexity Science and Engineering, The University of Tokyo()
2nd Author's Name Shotaro AKAHO
2nd Author's Affiliation The National Institute of Advanced Industrial Science and Technology (AIST) Neuroscience Research Institute
3rd Author's Name Masato OKADA
3rd Author's Affiliation Department of Complexity Science and Engineering, The University of Tokyo
Date 2007-10-18
Paper # NC2007-35
Volume (vol) vol.107
Number (no) 263
Page pp.pp.-
#Pages 5
Date of Issue