Bayesian principal component analysis with mixture priors
- Authors
- Oh, Hyun Sook; Kim, Dai-Gyoung
- Issue Date
- Sep-2010
- Publisher
- 한국통계학회
- Keywords
- Probabilistic principal component analysis; Dimension reduction; Probabilistic latent variable model
- Citation
- Journal of the Korean Statistical Society, v.39, no.3, pp.387 - 396
- Indexed
- SCIE
SCOPUS
KCI
- Journal Title
- Journal of the Korean Statistical Society
- Volume
- 39
- Number
- 3
- Start Page
- 387
- End Page
- 396
- URI
- https://scholarworks.bwise.kr/erica/handle/2021.sw.erica/39576
- DOI
- 10.1016/j.jkss.2010.04.001
- ISSN
- 1226-3192
- Abstract
- A central issue in principal component analysis (PCA) is that of choosing the appropriate number of principal components to be retained. Bishop (1999a) suggested a Bayesian approach for PCA for determining the effective dimensionality automatically on the basis of the probabilistic latent variable model. This paper extends this approach by using mixture priors, in that the choice dimensionality and estimation of principal components are done simultaneously via MCMC algorithm. Also, the proposed method provides a probabilistic measure of uncertainty on PCA, yielding posterior probabilities of all possible cases of principal components. (C) 2010 The Korean Statistical Society. Published by Elsevier B.V. All rights reserved.
- Files in This Item
-
Go to Link
- Appears in
Collections - COLLEGE OF SCIENCE AND CONVERGENCE TECHNOLOGY > ERICA 수리데이터사이언스학과 > 1. Journal Articles
Items in ScholarWorks are protected by copyright, with all rights reserved, unless otherwise indicated.