Generalized Canonical Correlation Analysis with Linear Constraints
A method is introduced by imposing linear constraints upon parameters corresponding to more than two sets of variables. We call the method introduced here ‘generalized canonical correlation analysis with linear constraint’. It covers canonical correlation analysis with linear constraints proposed by Yanai & Takane (1992) as its special case. Further, by employing dummy variables, our method turns out to be identical to the multiple correspondence analysis with linear constraints.
KeywordsWeight Vector Linear Constraint Canonical Correlation Canonical Correlation Analysis Multiple Correspondence Analysis
Unable to display preview. Download preview PDF.
- Bokenholt, U. Bockenholt,I. (1990): Canonical analysis of contingency tables with linear constraints, Psychometrika, 55, 633–639Google Scholar