Advances and Applications in Statistics
Volume 34, Issue 1, Pages 11 - 38
(May 2013)
|
|
CANONICAL CORRELATION ANALYSIS BETWEEN TWO VERTICAL MULTI-BLOCK TABLES AND APPLICATIONS TO QUALITATIVE VARIABLES
Gabriel Kissita, Léonard Niere, Roger Armand Makany and Dominique Mizere
|
Abstract: The canonical correlation analysis is introduced for the purpose of studying the link between two groups of quantitative variables measured on the same individual groups. It presents a minor practical interest, its theoretical basement is incontestable, because of it, we can obtain the factorial analysis of correspondence and discriminant analysis and, it is also the generalization of the linear regression. It can also be extended to several ways, we quote: general canonical correlation analysis (GCA), discriminant table analysis (DTA), multiple co-inertia analysis (MCIA), etc. The generalization of canonical correlation analysis hereby is based on researching the link between two multi-vertical tables as it accounts for STATICO in the simultaneous analysis of M couple of the tables. Additionally, this method is applied to qualitative variables. |
Keywords and phrases: data analysis, common components, specific weight, M-tables, STATIS, co-inertia, coupling of tables, STATICO. |
|
Number of Downloads: 358 | Number of Views: 1284 |
|