Distribution of the residual roots in principal components analysis.

  • A. M. Kshirsagar Defence Science Laboratory, Delhi
DOI: https://doi.org/10.14429/dsj.14.7162
Keywords: Covariance matrix, Residual roots

Abstract

The latent of distribution of latent roots of the covariance martix of normal variables, when a hypothetical linear function of the variables is eliminated, is derived in this paper. The relation between original roots and the residual roots- after elimination of, is also derived by an analytical method. An exact test for the goodness of fit of a single nonisotropic hypothetical principal components, using the residual roots, is then obtained.
Published
2014-05-09
How to Cite
Kshirsagar, A. (2014). Distribution of the residual roots in principal components analysis. Defence Science Journal, 14(4), 287-295. https://doi.org/10.14429/dsj.14.7162
Issue
Vol 14 No 4
Section
Research Papers
Copyright (c) 2016 Defence Science Journal

 Where otherwise noted, the Articles on this site are licensed under Creative Commons License: CC Attribution-Noncommercial-No Derivative Works 2.5 India