Improving principal component analysis using Bayesian estimation

Author

Nounou, Mohamed N. ; Bakshi, Bhavik R. ; Goel, Prem K. ; Shen, Xiaotong

Author_Institution

Dept. of Chem. Eng., Ohio State Univ., Columbus, OH, USA

Volume

5

fYear

2001

fDate

2001

Firstpage

3666

Abstract

Bayesian estimation is used in this paper to derive a new PCA (principal component analysis) modeling algorithm that improves the estimation accuracy by incorporating prior knowledge about the data and model. It is shown that the algorithm is more general than the existing methods [PCA and MLPCA (maximum-likelihood PCA)], and reduces to these techniques when a uniform prior is used. It is also shown that, when no external information is available, an empirically estimated prior from the available data can still provide improved accuracy over non-Bayesian methods

Keywords

Bayes methods; estimation theory; principal component analysis; Bayesian estimation; PCA modeling algorithm; empirically estimated prior; estimation accuracy; maximum-likelihood PCA; principal component analysis; prior knowledge; Additive noise; Bayesian methods; Chemical engineering; Density measurement; Matrix decomposition; Maximum likelihood estimation; Noise reduction; Pollution measurement; Principal component analysis; State estimation;

fLanguage

English

Publisher

ieee

Conference_Titel

American Control Conference, 2001. Proceedings of the 2001

Conference_Location

Arlington, VA

ISSN

0743-1619

Print_ISBN

0-7803-6495-3

Type

conf

DOI

10.1109/ACC.2001.946204

Filename

946204