DMCA
Canonical Ridge Analysis with Ridge Parameter Optimization (1998)
| Citations: | 7 - 3 self |
BibTeX
@MISC{Nielsen98canonicalridge,
author = {Finn Årup Nielsen and Lars Kai Hansen and F. A. Nielsen and L. K. Hansen and Stephen C. Strother},
title = {Canonical Ridge Analysis with Ridge Parameter Optimization},
year = {1998}
}
Share
 |
 |
 |
 |
||
OpenURL
Abstract
Introduction A number of linear multivariate statistical models have been applied for the analysis of functional images. We will here discuss a general model that encompasses several such approaches: Canonical Correlation Analysis (CCA), Partial Least Square (PLS), Orthonormalized Partial Least Square (OPLS), Ridge Regression and Fisher's Linear Discriminant (FLD). This general model is based on the concept of Canonical Ridge analysis [1]. The key matrix in this model is the "covariance" matrix: K = \Gamma (1 \Gamma kG ) G T G+ kG I \Delta \Gamma1=2 G T X \Gamma (1 \Gamma kX ) X T X+ kX I \Delta \Gamma1=2 = UV T (1) where the X is th
Keyphrases
canonical ridge analysis ridge parameter optimization delta gamma1 general model partial least square key matrix linear multivariate statistical model gamma kg several approach linear discriminant ridge regression canonical correlation analysis covariance matrix gamma kx functional image orthonormalized partial least square
