Page 507 - Jolliffe I. Principal Component Analysis

Index
472
relationships (QSAR), see
chemistry
least squares estimation,
multivariate regression, PC
quartimin/quartimax rotation 153, latent root regression,
154, 162–165, 270, 271, 277, regression, point estimation,
278 reduced rank regression,
quaternion valued data 370 ridge regression, robust
regression, selection of
ranked data 267, 338, 340, 341, variables
348, 349, 388 regression components 403
rank correlation 341 regression tree 185
reciprocal averaging, see scaling or regularized discriminant analysis
ordination 205, 207, 208
red noise 301, 304, 307, 314 reification 269
reduced rank regression 229, 230, repeatability of PCA 261, 394
331, 353, 392, 401 repeated measures, see longitudinal
softly shrunk reduced rank data
regression 230 rescaled PCs 403, 404
reduction of dimensionality, see residual variation 16, 17, 108, 114,
dimensionality reduction 129, 220, 240, 290, 399
redundancy analysis 225–230, 331, see also error covariance matrix,
393, 401 PCA of residuals
redundancy coefficient/index 226, residuals in a contingency table,
227 see interactions
redundant variables, see response variables 227–230
dimensionality reduction PCs of predicted responses 228,
regionalization studies 213, 294 230
regression analysis 13, 32, 33, 74, see also regression analysis
111, 112, 121, 127, 129, 137, restricted PC regression 184
144, 145, 157, 167–199, 202, ridge regression 167, 178, 179, 181,
205, 223, 227, 239, 240, 284, 185, 190, 364
286, 288, 290, 294, 304, 326, road running, see athletics
337, 352, 363, 366, 368, 378, robust estimation 232, 262–268
390, 399, 412 in functional PCA 266, 316, 327
computation 46, 168, 170, 173, in non-linear PCA 376
182, 412 in regression 264, 366
influence function 249, 250 of biplots 102, 265
interpretation 46, 168, 170, 173, of covariance/correlation
182, 412 matrices 264, 265–267, 363,
residuals 127, 399 364, 394
variable selection 111, 112, 137, of distributions of PCs 267
145, 167, 172, 182, 185–188, of means 241, 264, 265
190, 191, 194, 197, 198, 286 of PCs 50, 61, 233, 235, 263–268,
see also biased regression 356, 366, 368, 394, 401
methods, econometrics, of scale 266
influence functions, see also M-estimators, minimum