Canonical Concordance Correlation Analysis

A multivariate technique named Canonical Concordance Correlation Analysis (CCCA) is introduced. In contrast to the classical (CCA) which based on maximization of Pearson’s correlation coefficient between linear combinations two sets variables, CCCA maximizes Lin’s concordance accounts not just for maximum but also closeness aggregates’ mean values and their variances. While CCA employs centered data with excluded means can be understood as a more comprehensive characteristic similarity, or agreement measured simultaneously by distance variances, together possible aggregates variables in sets. The expressed generalized eigenproblem reduces regular if are equal, different it yields from solution. properties applications this type analysis described. approach useful solving various applied statistical problems when aggregated canonical correlations needed general