Generalizable patterns in neuroimaging:: How many principal components?

被引：131

作者：

Hansen, LK

Larsen, J

Nielsen, FÅ

Strother, SC

Rostrup, E

Savoy, R

Lange, N

Sidtis, J

Svarer, C

Paulson, OB

机构：

[1] Tech Univ Denmark, Dept Math Modeling, DK-2800 Lyngby, Denmark

[2] Univ Minnesota, Dept Neurol, Minneapolis, MN 55455 USA

[3] Univ Minnesota, Dept Radiol, Minneapolis, MN 55455 USA

[4] Univ Minnesota, Vet Affairs Med Ctr, PET Imaging Serv, Minneapolis, MN 55455 USA

[5] Univ Copenhagen, Hvidovre Hosp, Danish Ctr Magnet Resonance, DK-2650 Hvidovre, Denmark

[6] Massachusetts Gen Hosp, Dept Radiol, Charlestown, MA 02139 USA

[7] Harvard Univ, Sch Med, Belmont, MA 02178 USA

[8] McLean Hosp, Belmont, MA 02178 USA

[9] Rigshosp, Neurobiol Res Unit, DK-2100 Copenhagen, Denmark

来源：

NEUROIMAGE | 1999年 / 9卷 / 05期

关键词：

D O I：

10.1006/nimg.1998.0425

中图分类号：

Q189 [神经科学];

学科分类号：

071006 ;

摘要：

Generalization can be defined quantitatively and can be used to assess the performance of principal component analysis (PCA). The generalizability of PCA depends on the number of principal components retained in the analysis. We provide analytic and test set estimates of generalization. We show how the generalization error can be used to select the number of principal components in two analyses of functional magnetic resonance imaging activation sets. (C) 1999 Academic Press.

引用

页码：534 / 544

页数：11

共 30 条

[1] FITTING AUTOREGRESSIVE MODELS FOR PREDICTION [J].