|1.||Reducing dimensionality. If some variables are correlated, then the effective dimensionality may be decreased by expressing their information content in their PCs. If the first few PCs account for most of the variation in the original data, then the rest of the PCs may be neglected, reducing the effective dimensionality of the problem.|
|2.||Arranging variables into meaningful groups. It may be possible to interpret or assign meaning to the set of variables contributing significantly to the more important PCs. The PCs may also be rotated to maximize the difference in loadings between variables. Rotation modifies the PCs without altering their underlying structure, to allow easier determination of the significant variables in each PC and better interpretation.|
|3.||Discarding redundant variables. Some variables may be redundant and may therefore be discarded. PCA gives a method for systematically identifying variables with the least contribution to the overall variability in the data.|
Questia, a part of Gale, Cengage Learning. www.questia.com
Publication information: Book title: Global Environmental Risk. Contributors: Jeanne X. Kasperson - Editor, Roger Kasperson - Editor. Publisher: United Nations University Press. Place of publication: New York. Publication year: 2001. Page number: 140.
This material is protected by copyright and, with the exception of fair use, may not be further copied, distributed or transmitted in any form or by any means.