Page 494 - Jolliffe I. Principal Component Analysis
P. 494
459
Index
116, 127-130, 133, 270, 274,
284, 289, 294–297, 302–317,
343, 346
presence/absence data 105, 107,
332, 354, 362, 364, 385 binary variables 68, 88, 339, 340,
see also meteorology and 390
climatology biological applications 9, 57, 64,
autocorrelation/autocovariance 90, 390
129, 149, 298, 335, 336 aphids 122, 145-147, 214, 219
autoregressive processes birds
ARMA models 336, 337 Irish wetland birds 105–106
first order 301, 327, 334, 335 seabird communities 214
multivariate first order 302, 308 diving seals 316, 323
auxiliaries 399 see also ecology, size and shape
PCs
Bartlett’s test see hypothesis biplots 79, 90–103, 132, 230, 342,
testing for equality of PC 353, 408
variances bimodel 91
basis functions 318–320, 325, 327 classical 90, 101
Fourier basis 319 coefficient of variation 102, 389
spline basis 320, 322, 325, 331 computation 413
Bayesian inference 222, 395 correspondence analysis 95, 96
in factor analysis 155 generalized 102
in regression 177, 179 interpolation and prediction 102,
posterior distribution 56, 126 382
prior distribution 60, 126, 179 non-linear 102, 381, 382
using PCs 56, 60 robust 102, 265
Behrens-Fisher problem, symmetric 96
multivariate 356 bisection method 411
best-fitting Bonferroni bound 248
lines, planes and hyperplanes 7, bootstrap estimation 49, 112, 117,
36, 189, 389 125, 126, 261, 267, 314, 394
subspaces 34–36, 78, 80, 87, 342 confidence intervals 52, 118, 126,
best linear approximation to PCs 331
using subset of variables in quality control 368
294 non-parametric 52
best linear predictors 16, 17, 392 of residuals 125, 377
between–group variation 202, 203, parametric 52
209, 220, 399 resampling in blocks 360
between–treatment (and –block) boxplots 295
PCs 352 of PC scores 125, 235
biased regression methods 167, branch-and-bound algorithm 284
168, 171, 172, 177–179, broken stick model 115, 130, 132,
183–185, 230, 286 143
see also PC regression, ridge Burt matrix 343
regression, shrinkage
methods calibration 190

