Multivariate Reduced-Rank Regression
摘要
Multivariate reduced-rank regression is a way of constraining the multivariate linear regression model so that the rank of the regression coefficient matrix has less than full rank. Without the constraint, multivariate linear regression has no true multivariate content.