Comparison Of Canonical Correlation Analysis And The Generalized Canonical Correlation Analysis Using The Lognormal And Cauchy Distributed Data

Abstract

This work compares the Canonical Correlation Analysis and Generalized Canonical Correlation Analysis for two data sets using Lognormal and Cauchy distributions. Data of different sample sizes (10, 20, 30, 40 and 50) were simulated and used to perform the analysis. Their relative efficiencies were calculated and it was discovered to be the same for two data sets except for scaling.

Keywords: Relative efficiency, Scaling, Canonical Correlation Analysis, Generalized Canonical Correlation Analysis, Lognormal, Cauchy