A Dual Method of Carroll's Generalized Canonical Correlation Analysis
Rodnellin Onesime Malouata *
Departement of Licence, Higher Institute of Management, Congo.
Gelin Chedly Louzayadio
Departement of Licence, Faculty of Economics, Congo.
Michel Koukouatikissa Diafouka
Departement of Licence, Teachers' College, Congo.
Leonard Niere
Departement of Licence, Higher Institute of Management, Congo.
*Author to whom correspondence should be addressed.
Abstract
Aims/Objectives: We propose a dual method of Carroll's generalized canonical correlation analysis and we prove by means of the proposed criterion that the duality is formulated by exchanging the operators. It is an extension of Carroll's generalized canonical correlation analysis. The approach of analysis is illustrated on the basis of case study.
Keywords: Dual generalized canonical correlation analysis, data set partitioned in groups of individuals, principal component analysis