A variational bayesian approach for unsupervised clustering

Mu Song Chen, Hsuan Fu Wang, Chipan Hwang, Tze Yee Ho, Chan Hsiang Hung

Research output: Chapter in Book/Report/Conference proceedingChapter

1 Citation (Scopus)

Abstract

Gaussian Mixture Models are among the most statistically mature methods which are used to make statistical inferences as well as performing unsupervised clustering. Formally, a gaussian mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the data set. In this paper, a probabilistic clustering based on the finite mixture models of the data distribution is suggested. An important issue in the finite mixture model-based clustering approach is to select the number of mixture components of clusters. In this sense, we focus on statistical inference for finite mixture models and illustrate how the variational Bayesian approach can be used to determine a suitable number of components in the case of a mixture of Gaussian distributions.

Original languageEnglish
Title of host publicationLecture Notes in Electrical Engineering
PublisherSpringer Verlag
Pages651-660
Number of pages10
DOIs
Publication statusPublished - 2016 Jan 1

Publication series

NameLecture Notes in Electrical Engineering
Volume375
ISSN (Print)1876-1100
ISSN (Electronic)1876-1119

Fingerprint

Gaussian distribution
Probability distributions

All Science Journal Classification (ASJC) codes

  • Industrial and Manufacturing Engineering

Cite this

Chen, M. S., Wang, H. F., Hwang, C., Ho, T. Y., & Hung, C. H. (2016). A variational bayesian approach for unsupervised clustering. In Lecture Notes in Electrical Engineering (pp. 651-660). (Lecture Notes in Electrical Engineering; Vol. 375). Springer Verlag. https://doi.org/10.1007/978-981-10-0539-8_63
Chen, Mu Song ; Wang, Hsuan Fu ; Hwang, Chipan ; Ho, Tze Yee ; Hung, Chan Hsiang. / A variational bayesian approach for unsupervised clustering. Lecture Notes in Electrical Engineering. Springer Verlag, 2016. pp. 651-660 (Lecture Notes in Electrical Engineering).
@inbook{54d5d2cedb074741bf0f647ce3cf5b22,
title = "A variational bayesian approach for unsupervised clustering",
abstract = "Gaussian Mixture Models are among the most statistically mature methods which are used to make statistical inferences as well as performing unsupervised clustering. Formally, a gaussian mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the data set. In this paper, a probabilistic clustering based on the finite mixture models of the data distribution is suggested. An important issue in the finite mixture model-based clustering approach is to select the number of mixture components of clusters. In this sense, we focus on statistical inference for finite mixture models and illustrate how the variational Bayesian approach can be used to determine a suitable number of components in the case of a mixture of Gaussian distributions.",
author = "Chen, {Mu Song} and Wang, {Hsuan Fu} and Chipan Hwang and Ho, {Tze Yee} and Hung, {Chan Hsiang}",
year = "2016",
month = "1",
day = "1",
doi = "10.1007/978-981-10-0539-8_63",
language = "English",
series = "Lecture Notes in Electrical Engineering",
publisher = "Springer Verlag",
pages = "651--660",
booktitle = "Lecture Notes in Electrical Engineering",
address = "Germany",

}

Chen, MS, Wang, HF, Hwang, C, Ho, TY & Hung, CH 2016, A variational bayesian approach for unsupervised clustering. in Lecture Notes in Electrical Engineering. Lecture Notes in Electrical Engineering, vol. 375, Springer Verlag, pp. 651-660. https://doi.org/10.1007/978-981-10-0539-8_63

A variational bayesian approach for unsupervised clustering. / Chen, Mu Song; Wang, Hsuan Fu; Hwang, Chipan; Ho, Tze Yee; Hung, Chan Hsiang.

Lecture Notes in Electrical Engineering. Springer Verlag, 2016. p. 651-660 (Lecture Notes in Electrical Engineering; Vol. 375).

Research output: Chapter in Book/Report/Conference proceedingChapter

TY - CHAP

T1 - A variational bayesian approach for unsupervised clustering

AU - Chen, Mu Song

AU - Wang, Hsuan Fu

AU - Hwang, Chipan

AU - Ho, Tze Yee

AU - Hung, Chan Hsiang

PY - 2016/1/1

Y1 - 2016/1/1

N2 - Gaussian Mixture Models are among the most statistically mature methods which are used to make statistical inferences as well as performing unsupervised clustering. Formally, a gaussian mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the data set. In this paper, a probabilistic clustering based on the finite mixture models of the data distribution is suggested. An important issue in the finite mixture model-based clustering approach is to select the number of mixture components of clusters. In this sense, we focus on statistical inference for finite mixture models and illustrate how the variational Bayesian approach can be used to determine a suitable number of components in the case of a mixture of Gaussian distributions.

AB - Gaussian Mixture Models are among the most statistically mature methods which are used to make statistical inferences as well as performing unsupervised clustering. Formally, a gaussian mixture model corresponds to the mixture distribution that represents the probability distribution of observations in the data set. In this paper, a probabilistic clustering based on the finite mixture models of the data distribution is suggested. An important issue in the finite mixture model-based clustering approach is to select the number of mixture components of clusters. In this sense, we focus on statistical inference for finite mixture models and illustrate how the variational Bayesian approach can be used to determine a suitable number of components in the case of a mixture of Gaussian distributions.

UR - http://www.scopus.com/inward/record.url?scp=85007571997&partnerID=8YFLogxK

UR - http://www.scopus.com/inward/citedby.url?scp=85007571997&partnerID=8YFLogxK

U2 - 10.1007/978-981-10-0539-8_63

DO - 10.1007/978-981-10-0539-8_63

M3 - Chapter

T3 - Lecture Notes in Electrical Engineering

SP - 651

EP - 660

BT - Lecture Notes in Electrical Engineering

PB - Springer Verlag

ER -

Chen MS, Wang HF, Hwang C, Ho TY, Hung CH. A variational bayesian approach for unsupervised clustering. In Lecture Notes in Electrical Engineering. Springer Verlag. 2016. p. 651-660. (Lecture Notes in Electrical Engineering). https://doi.org/10.1007/978-981-10-0539-8_63