Robust Bilinear Probabilistic Principal Component Analysis

Principal component analysis (PCA) is one of the most popular tools in multivariate exploratory data analysis. Its probabilistic version (PPCA) based on the maximum likelihood procedure provides a probabilistic manner to implement dimension reduction. Recently, the bilinear PPCA (BPPCA) model, which...

Descripción completa

Guardado en:
Detalles Bibliográficos
Autores principales: Yaohang Lu, Zhongming Teng
Formato: article
Lenguaje:EN
Publicado: MDPI AG 2021
Materias:
Acceso en línea:https://doaj.org/article/a226d30d8ec741a4a23804bedfa1a54b
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
id oai:doaj.org-article:a226d30d8ec741a4a23804bedfa1a54b
record_format dspace
spelling oai:doaj.org-article:a226d30d8ec741a4a23804bedfa1a54b2021-11-25T16:13:10ZRobust Bilinear Probabilistic Principal Component Analysis10.3390/a141103221999-4893https://doaj.org/article/a226d30d8ec741a4a23804bedfa1a54b2021-11-01T00:00:00Zhttps://www.mdpi.com/1999-4893/14/11/322https://doaj.org/toc/1999-4893Principal component analysis (PCA) is one of the most popular tools in multivariate exploratory data analysis. Its probabilistic version (PPCA) based on the maximum likelihood procedure provides a probabilistic manner to implement dimension reduction. Recently, the bilinear PPCA (BPPCA) model, which assumes that the noise terms follow matrix variate Gaussian distributions, has been introduced to directly deal with two-dimensional (2-D) data for preserving the matrix structure of 2-D data, such as images, and avoiding the curse of dimensionality. However, Gaussian distributions are not always available in real-life applications which may contain outliers within data sets. In order to make BPPCA robust for outliers, in this paper, we propose a robust BPPCA model under the assumption of matrix variate <i>t</i> distributions for the noise terms. The alternating expectation conditional maximization (AECM) algorithm is used to estimate the model parameters. Numerical examples on several synthetic and publicly available data sets are presented to demonstrate the superiority of our proposed model in feature extraction, classification and outlier detection.Yaohang LuZhongming TengMDPI AGarticle2-D dataprobabilistic principal component analysisAECM algorithmmatrix variate Gaussian distributionsmatrix variate <i>t</i> distributionsoutliersIndustrial engineering. Management engineeringT55.4-60.8Electronic computers. Computer scienceQA75.5-76.95ENAlgorithms, Vol 14, Iss 322, p 322 (2021)
institution DOAJ
collection DOAJ
language EN
topic 2-D data
probabilistic principal component analysis
AECM algorithm
matrix variate Gaussian distributions
matrix variate <i>t</i> distributions
outliers
Industrial engineering. Management engineering
T55.4-60.8
Electronic computers. Computer science
QA75.5-76.95
spellingShingle 2-D data
probabilistic principal component analysis
AECM algorithm
matrix variate Gaussian distributions
matrix variate <i>t</i> distributions
outliers
Industrial engineering. Management engineering
T55.4-60.8
Electronic computers. Computer science
QA75.5-76.95
Yaohang Lu
Zhongming Teng
Robust Bilinear Probabilistic Principal Component Analysis
description Principal component analysis (PCA) is one of the most popular tools in multivariate exploratory data analysis. Its probabilistic version (PPCA) based on the maximum likelihood procedure provides a probabilistic manner to implement dimension reduction. Recently, the bilinear PPCA (BPPCA) model, which assumes that the noise terms follow matrix variate Gaussian distributions, has been introduced to directly deal with two-dimensional (2-D) data for preserving the matrix structure of 2-D data, such as images, and avoiding the curse of dimensionality. However, Gaussian distributions are not always available in real-life applications which may contain outliers within data sets. In order to make BPPCA robust for outliers, in this paper, we propose a robust BPPCA model under the assumption of matrix variate <i>t</i> distributions for the noise terms. The alternating expectation conditional maximization (AECM) algorithm is used to estimate the model parameters. Numerical examples on several synthetic and publicly available data sets are presented to demonstrate the superiority of our proposed model in feature extraction, classification and outlier detection.
format article
author Yaohang Lu
Zhongming Teng
author_facet Yaohang Lu
Zhongming Teng
author_sort Yaohang Lu
title Robust Bilinear Probabilistic Principal Component Analysis
title_short Robust Bilinear Probabilistic Principal Component Analysis
title_full Robust Bilinear Probabilistic Principal Component Analysis
title_fullStr Robust Bilinear Probabilistic Principal Component Analysis
title_full_unstemmed Robust Bilinear Probabilistic Principal Component Analysis
title_sort robust bilinear probabilistic principal component analysis
publisher MDPI AG
publishDate 2021
url https://doaj.org/article/a226d30d8ec741a4a23804bedfa1a54b
work_keys_str_mv AT yaohanglu robustbilinearprobabilisticprincipalcomponentanalysis
AT zhongmingteng robustbilinearprobabilisticprincipalcomponentanalysis
_version_ 1718413272790597632