Robust Methods for Canonical Correlation Analysis
نویسندگان
چکیده
Canonical correlation analysis studies associations between two sets of random variables. Its standard computation is based on sample covariance matrices, which are however very sensitive to outlying observations. In this note we introduce, discuss and compare four different ways for performing a robust canonical correlation analysis. One method uses robust estimators of the involved covariance matrices, another one uses the signs of the observations, a third approach is based on projection pursuit, and finally an alternating regression algorithm for canonical analysis is proposed.
منابع مشابه
Robust Canonical Correlations: a Comparative Study
Several approaches for robust canonical correlation analysis will be presented and discussed. A first method is based on the definition of canonical correlation analysis as looking for linear combinations of two sets of variables having maximal (robust) correlation. A second method is based on alternating robust regressions. These methods are discussed in detail and compared with the more tradi...
متن کاملCanonical Correlation Analysis for Determination of Relationship between Morphological and Physiological Pollinated Characteristics in Five Varieties of Phalaenopsis
Phalaenopsis is an important genus of orchids that is grown for economical production of cut flower and potted plants. The objective of this study is the evaluation of correlation between morphological and physiological traits of self and cross-pollination of 5 varieties of Phalaenopsis orchid. Some morphological traits were measured: Capsule length (CL), capsule volume (CV), weight of seeds in...
متن کاملCanonical Analysis of the Relationship between Components of Professional Ethics and Dimensions of Social Responsibility
Background: Today, professional ethics and social responsibility play an important role in organizations. This study aimed canonical analysis of the relationship between components of professional ethics and social responsibility dimensions among the first high school teachers in the Naghadeh province. Method: This study, in terms of purpose is application, and in terms of data collec...
متن کاملGeneralization of Canonical Correlation Analysis from Multivariate to Functional Cases and its related problems
In multivariate cases, the aim of canonical correlation analysis (CCA) for two sets of variables x and y is to obtain linear combinations of them so that they have the largest possible correlation. However, when x and y are continouse functions of another variable (generally time) in nature, these two functions belong to function spaces which are of infinite dimension, and CCA for them should b...
متن کاملGene-Gene association for Imaging Genetics Data using Robust Kernel Canonical Correlation Analysis
In genome-wide interaction studies, to detect gene-gene interactions, most methods are divided into two folds: single nucleotide polymorphisms (SNP) based and gene-based methods. Basically, the methods based on the gene are more effective than the methods based on a single SNP. Recent years, while the kernel canonical correlation analysis (Classical kernel CCA) based U statistic (KCCU) has prop...
متن کامل