Canonical analysis of several sets of variables

作者：Jon R. Kettenring · 发表于：Biometrika · 年份：1971 · DOI：10.1093/biomet/58.3.433 · 被引用次数：795 · 研究领域：Advanced Statistical Methods and Models、Optimal Experimental Design Methods、Spectroscopy and Chemometric Analyses

Five extensions of the classical two-set theory of canonical correlation analysis to three or more sets are considered. For each one, a model of the general principal component type is constructed to aid in motivating, comparing and understanding the methods. Procedures are developed for finding the canonical variables associated with the different approaches. Some practical considerations and an example are also included.