Abstract
This paper investigates the problem of the choice of dimension in Principal Component Analysis (PCA). PCA is introduced as a model; a loss function assessing the stability of the fit is considered. The choice of dimension then amounts to the minimisation of an expected loss which has to be estimated. This is achieved by resampling methods. Different bootstrap and jackknife estimates are presented. The behaviour of these estimates are investigated on artificial data and on real data. The resulting choices are confronted with those given by naïve rules.
This is a preview of subscription content, log in via an institution to check access.
Access this chapter
Tax calculation will be finalised at checkout
Purchases are for personal use only
Preview
Unable to display preview. Download preview PDF.
References
Becker, R.A., Chambers, J.M., Wilks, A.R. (1988). The New S Language, a Programming Environment for Data Analysis and Graphics, Wadsworth and Brooks/Cole, Pacific Grove, Ca 93950.
Beran, R., Srivastava, M.S. (1985). Bootstrap Tests and Confidence Regions for Functions of a Covariance Matrix. The Annals of Statistics, 13, 95–115.
Besse, Ph. (1992). PCA Stability and Choice of Dimensionality. Statistics hi Probability Letters, 13, 405–410.
Besse, P., Caussinus, H., Ferré, L., Fine, J. (1988). Principal Components Analysis and Optimization of Graphical Displays. Statistics, 19, 301–312.
Besse, Ph., Pousse, A. (1992). Extension des Analyses Factorielles, in Modèles pour l’Analyse des Données Multidimensionnelles, J.J. Droesbeke et al. (eds.), Economica, Paris.
Caussinus, H. (1986). Models and Uses of Principal Component Analysis, in Multidimensional Data Analysis, J. de Leeuw et al. (eds.), DSWO Press, Leiden, 149–170.
Daudin, J.J., Duby, C., Trécourt, P. (1988). Stability of Principal Component Analysis Studied by Bootstrap, Statistics, 19, 241–158.
Daudin, J.J., Duby, C., Trécourt, P. (1989). PCA Stability Studied by the Bootstrap and the Infinitesimal Jackknife Method, Statistics, 20, 255–270.
Efron, B. (1982). The Jackknife, the Bootstrap and other Resampling Methods, SIAM, Philadelphie.
Efron, B. (1992). Jackknife-after-Bootstrap Standard Errors and Influence Functions (with discussion), Journal of the Royal Statistical Society, series B, 54, 83–127.
Fine, J., Pousse, A. (1991). Asymptotic Study of the Multivariate Functional Model; Application to the Metric Choice in PCA, Statistics, to appear.
Jolliffe, I. (1986). Principal Component Analysis, Springer-Verlag, New-York.
Kato, T. (1966). Perturbation Theory for Linear Operator, Springer-Verlag, New-York. McDonald, G.C., Schwing, R.C. (1973). Instabilities of Regression Estimates Relating Air Pollution to Mortality. Technometrics, 15, 463–481.
SAS (1989), SAS/STAT User’s Guide, volume 2, Version 6, fourth edition, Sas Institute Inc, Cary.
Author information
Authors and Affiliations
Editor information
Editors and Affiliations
Rights and permissions
Copyright information
© 1993 Springer-Verlag Berlin Heidelberg
About this paper
Cite this paper
Besse, P., de Falguerolles, A. (1993). Application of Resampling Methods to the Choice of Dimension in Principal Component Analysis. In: Härdle, W., Simar, L. (eds) Computer Intensive Methods in Statistics. Statistics and Computing. Physica, Heidelberg. https://doi.org/10.1007/978-3-642-52468-4_11
Download citation
DOI: https://doi.org/10.1007/978-3-642-52468-4_11
Publisher Name: Physica, Heidelberg
Print ISBN: 978-3-7908-0677-9
Online ISBN: 978-3-642-52468-4
eBook Packages: Springer Book Archive