Weighted sparse principal component analysis

Katrijn Van Deun*, Lieven Thorrez, Marguerita Coccia, Dicle Hasdemir, Johan Westerhuis, A.K. Smilde, Iven Van Mechelen

*Corresponding author for this work

Research output: Contribution to journalArticleScientificpeer-review

Abstract

Sparse principal component analysis (SPCA) has been shown to be a fruitful method for the analysis of high-dimensional data. So far, however, no method has been proposed that allows to assign elementwise weights to the matrix of residuals, although this may have several useful applications. We propose a novel SPCA method that includes the flexibility to weight at the level of the elements of the data matrix. The superior performance of the weighted SPCA approach compared to unweighted SPCA is shown for data simulated according to the prevailing multiplicative-additive error model. In addition, applying weighted SPCA to genomewide transcription rates obtained soon after vaccination, resulted in a biologically meaningful selection of variables with components that are associated to the measured vaccine efficacy. The MATLAB implementation of the weighted sparse PCA method is freely available from https://github.com/katrijnvandeun/WSPCA.
Original languageEnglish
Article number103875
JournalChemometrics & Intelligent Laboratory Systems
Volume195
DOIs
Publication statusPublished - 2019

Fingerprint

Principal component analysis
Vaccines
Transcription
MATLAB

Keywords

  • DECOMPOSITION
  • Elementwise weighted least squares
  • LASSO
  • LEAST-SQUARES REGRESSION
  • METABOLOMICS
  • MODELS
  • Multiplicative-additive error
  • PCA
  • SELECTION
  • SHRINKAGE
  • Sparse principal component analysis

Cite this

Van Deun, K., Thorrez, L., Coccia, M., Hasdemir, D., Westerhuis, J., Smilde, A. K., & Van Mechelen, I. (2019). Weighted sparse principal component analysis. Chemometrics & Intelligent Laboratory Systems, 195, [103875]. https://doi.org/10.1016/j.chemolab.2019.103875
Van Deun, Katrijn ; Thorrez, Lieven ; Coccia, Marguerita ; Hasdemir, Dicle ; Westerhuis, Johan ; Smilde, A.K. ; Van Mechelen, Iven. / Weighted sparse principal component analysis. In: Chemometrics & Intelligent Laboratory Systems. 2019 ; Vol. 195.
@article{44bc2f645bcf41e6b67d86a7aea1c753,
title = "Weighted sparse principal component analysis",
abstract = "Sparse principal component analysis (SPCA) has been shown to be a fruitful method for the analysis of high-dimensional data. So far, however, no method has been proposed that allows to assign elementwise weights to the matrix of residuals, although this may have several useful applications. We propose a novel SPCA method that includes the flexibility to weight at the level of the elements of the data matrix. The superior performance of the weighted SPCA approach compared to unweighted SPCA is shown for data simulated according to the prevailing multiplicative-additive error model. In addition, applying weighted SPCA to genomewide transcription rates obtained soon after vaccination, resulted in a biologically meaningful selection of variables with components that are associated to the measured vaccine efficacy. The MATLAB implementation of the weighted sparse PCA method is freely available from https://github.com/katrijnvandeun/WSPCA.",
keywords = "DECOMPOSITION, Elementwise weighted least squares, LASSO, LEAST-SQUARES REGRESSION, METABOLOMICS, MODELS, Multiplicative-additive error, PCA, SELECTION, SHRINKAGE, Sparse principal component analysis",
author = "{Van Deun}, Katrijn and Lieven Thorrez and Marguerita Coccia and Dicle Hasdemir and Johan Westerhuis and A.K. Smilde and {Van Mechelen}, Iven",
year = "2019",
doi = "10.1016/j.chemolab.2019.103875",
language = "English",
volume = "195",
journal = "Chemometrics & Intelligent Laboratory Systems",
issn = "0169-7439",
publisher = "Elsevier Science BV",

}

Van Deun, K, Thorrez, L, Coccia, M, Hasdemir, D, Westerhuis, J, Smilde, AK & Van Mechelen, I 2019, 'Weighted sparse principal component analysis', Chemometrics & Intelligent Laboratory Systems, vol. 195, 103875. https://doi.org/10.1016/j.chemolab.2019.103875

Weighted sparse principal component analysis. / Van Deun, Katrijn; Thorrez, Lieven; Coccia, Marguerita; Hasdemir, Dicle; Westerhuis, Johan; Smilde, A.K.; Van Mechelen, Iven.

In: Chemometrics & Intelligent Laboratory Systems, Vol. 195, 103875, 2019.

Research output: Contribution to journalArticleScientificpeer-review

TY - JOUR

T1 - Weighted sparse principal component analysis

AU - Van Deun, Katrijn

AU - Thorrez, Lieven

AU - Coccia, Marguerita

AU - Hasdemir, Dicle

AU - Westerhuis, Johan

AU - Smilde, A.K.

AU - Van Mechelen, Iven

PY - 2019

Y1 - 2019

N2 - Sparse principal component analysis (SPCA) has been shown to be a fruitful method for the analysis of high-dimensional data. So far, however, no method has been proposed that allows to assign elementwise weights to the matrix of residuals, although this may have several useful applications. We propose a novel SPCA method that includes the flexibility to weight at the level of the elements of the data matrix. The superior performance of the weighted SPCA approach compared to unweighted SPCA is shown for data simulated according to the prevailing multiplicative-additive error model. In addition, applying weighted SPCA to genomewide transcription rates obtained soon after vaccination, resulted in a biologically meaningful selection of variables with components that are associated to the measured vaccine efficacy. The MATLAB implementation of the weighted sparse PCA method is freely available from https://github.com/katrijnvandeun/WSPCA.

AB - Sparse principal component analysis (SPCA) has been shown to be a fruitful method for the analysis of high-dimensional data. So far, however, no method has been proposed that allows to assign elementwise weights to the matrix of residuals, although this may have several useful applications. We propose a novel SPCA method that includes the flexibility to weight at the level of the elements of the data matrix. The superior performance of the weighted SPCA approach compared to unweighted SPCA is shown for data simulated according to the prevailing multiplicative-additive error model. In addition, applying weighted SPCA to genomewide transcription rates obtained soon after vaccination, resulted in a biologically meaningful selection of variables with components that are associated to the measured vaccine efficacy. The MATLAB implementation of the weighted sparse PCA method is freely available from https://github.com/katrijnvandeun/WSPCA.

KW - DECOMPOSITION

KW - Elementwise weighted least squares

KW - LASSO

KW - LEAST-SQUARES REGRESSION

KW - METABOLOMICS

KW - MODELS

KW - Multiplicative-additive error

KW - PCA

KW - SELECTION

KW - SHRINKAGE

KW - Sparse principal component analysis

U2 - 10.1016/j.chemolab.2019.103875

DO - 10.1016/j.chemolab.2019.103875

M3 - Article

VL - 195

JO - Chemometrics & Intelligent Laboratory Systems

JF - Chemometrics & Intelligent Laboratory Systems

SN - 0169-7439

M1 - 103875

ER -