The effects on the distributions of sample canonical correlations under nonnormality

In this paper, we study the effects of nonnormality on the distributions of sample canonical correlations when the population canonical correlations are simple. In order to achieve the purpose, we derive asymptotic expansion formulas for the distributions of a function of the canonical correlations as well as the individual canonical correlations under nonnormal populations. We particularly discuss the distribution of sample canonical correlations under the class of elliptical population. These expansions are given by using a perturbation method. Simulation results are also given.     

Interpretation of Canonical Discriminant Functions,Canonical Variates,and Principal Components

Canonical discriminant functions are defined here as linear combinations that separate groups of observations, and canonical variates are defined as linear combinations associated with canonical correlations between two sets of variables. In standardized form, the coefficients in either type of canonical function provide information about the joint contribution of the variables to the canonical function. The standardized coefficients can be converted to correlations between the variables and the canonical function. These correlations generally alter the interpretation of the canonical functions. For canonical discriminant functions, the standardized coefficients are compared with the correlations, with partial t and F tests, and with rotated coefficients. For canonical variates, the discussion includes standardized coefficients, correlations between variables and the function, rotation, and redundancy analysis. Various approaches to interpretation of principal components are compared: the choice between the covariance and correlation matrices, the conversion of coefficients to correlations, the rotation of the coefficients, and the effect of special patterns in the covariance and correlation matrices.     

Canonical Correlation Analysis Using Small Number of Samples

Canonical correlation assesses the relationship between two groups of variables. Although it has been a useful tool in a wide variety of research areas, it is not well known that weaker canonical correlations require larger sample sizes to be correctly inferred. In this article, we investigate small sample bias in canonical correlation analysis and apply the jackknife bias correction to the estimation of canonical correlations. We use bootstrap samples to obtain a better confidence interval for the jackknife canonical correlation estimator.     

some results on cononical correlations and measures of multivariate association

The canonical correlations and several orher measures of multivariate association between two sets of variables (x and y) are considered when the covariance matrices are singular. A useful inequality for the canonical correlations when new vari- ables are brought into x or y is obtained for both the nonsingular and singular cases. It is also shown that, under a simple condition, measures of multivariate association equal one if and only if there exists a linear relationship between the sets of variables.     

Parallel analysis approach for determining dimensionality in canonical correlation analysis

ABSTRACT

Canonical correlations are maximized correlation coefficients indicating the relationships between pairs of canonical variates that are linear combinations of the two sets of original variables. The number of non-zero canonical correlations in a population is called its dimensionality. Parallel analysis (PA) is an empirical method for determining the number of principal components or factors that should be retained in factor analysis. An example is given to illustrate for adapting proposed procedures based on PA and bootstrap modified PA to the context of canonical correlation analysis (CCA). The performances of the proposed procedures are evaluated in a simulation study by their comparison with traditional sequential test procedures with respect to the under-, correct- and over-determination of dimensionality in CCA.     

On multivariate linear regression shrinkage and reduced-rank procedures

An alternate derivation of the canonical analysis shrinkage prediction procedure of Breiman and Friedman (1997. J. Roy. Statist. Soc. B 59, 3–54) is presented for the multivariate linear model. It is based on consideration of prediction mean square error matrix, and bias of the squared sample canonical correlations. A modified procedure involving partial canonical correlation analysis is also introduced and discussed.     

Study of redundancy of vector variables in canonical correlations

Fujikoshi (1982) obtained the necessary and sufficient conditions for the increased number of variables in the two sets of vectors not affecting the original nonzero canonical correlations and used these to obtain the likelihood ratio test procedure. He assumed a nonsingular covariance matrix due to random variables. Here, we study the same problem when the covariance matrix is singular and establish some further results. In this study, we note that the unit canonical correlations have to be separated in some of the situations. These results are valid for complex random vector variables and in some situations, the test for redundancy is given for complex random variables.     

A Note on the Canonical Correlations From Contingency Tables

Explicit formulae are obtained for the asymptotic variances and covariances of canonical correlations which correspond to non-zero theoretical correlations in (p+ 1) x (q+1) contingency tables, with p≤q. The moments of the roots of a central Wishart matrix distributed as Wp(q; I ) are also given in general, with means, variances and covariances tabulated for p= 2, 3, 4: these may apply to canonical correlations corresponding to zeros.     

Canonical correlation analysis for the vector AR(1) model with ARCH innovations

This paper extends the results of canonical correlation analysis of Anderson [2002. Canonical correlation analysis and reduced-rank regression in autoregressive models. Ann. Statist. 30, 1134–1154] to a vector AR(1) process with a vector ARCH(1) innovations. We obtain the limiting distributions of the sample matrices, the canonical correlations and the canonical vectors of the process. The extension is important because many time series in economics and finance exhibit conditional heteroscedasticity. We also use simulation to demonstrate the effects of ARCH innovations on the canonical correlation analysis in finite sample. Both the limiting distributions and simulation results show that overlooking the ARCH effects in canonical correlation analysis can easily lead to erroneous inference.     

Two Canonical VARMA Forms: Scalar Component Models Vis-à-Vis the Echelon Form

In this article we study two methodologies which identify and specify canonical form VARMA models. The two methodologies are: (1) an extension of the scalar component methodology which specifies canonical VARMA models by identifying scalar components through canonical correlations analysis; and (2) the Echelon form methodology, which specifies canonical VARMA models through the estimation of Kronecker indices. We compare the actual forms and the methodologies on three levels. Firstly, we present a theoretical comparison. Secondly, we present a Monte Carlo simulation study that compares the performances of the two methodologies in identifying some pre-specified data generating processes. Lastly, we compare the out-of-sample forecast performance of the two forms when models are fitted to real macroeconomic data.     

A Comparison of Some Tests for Determining the Number of Nonzero Canonical Correlations

When considering the relationships between two sets of variates, the number of nonzero population canonical correlations may be called the dimensionality. In the literature, several tests for dimensionality in the canonical correlation analysis are known. A comparison of seven sequential test procedures is presented, using results from some simulation study. The tests are compared with regard to the relative frequencies of underestimation, correct estimation, and overestimation of the true dimensionality. Some conclusions from the simulation results are drawn.     

A generalization of random vectors and a correlation theory for them - a trial (with discussions) 2

There are defined generalized generalized random vectors. This notion contains usual random vectors, some classical and generalized stochastic processes and certain generalizations of them. There is developped a theory of correlation on basis of a theory of linear prediction (approximation) for such rendom vectors which includes especially canonical correlations.     

Multivariate regression analysis and canonical variates

A method called FICYREG of estimating regression coefficients is introduced. This is a generalization to the multivariate regression problem of the James-Stein estimator. When suitably représentés FICYREG emerges as a rule in which the canonical variates and canonical correlations have an intrinsic role to play. By exploiting these objects FICYREG is able to achieve stability against the influence of the “noise” present in problems where the responses are correlated so that some of the response vector's canonical variates will be essentially independent of all others including the predictors. The least squares (LS) estimator is, by contrast, highly sensitive to this noise. The use of FICYREG is illustrated in terms of an example, and its peformance is compared to the LS estimator when a quadratic loss function is assumed. The cases of both fixed and random predictors are considered. Overall, FICYREG outperforms the LS estimator.     

Stepwise canonical analysis in categorical data

Assignment of optimum scores to the categories of a multidimensional contingency table by use of a stepwise method of canonical correlations for more than two vector variables is proposed and illustrated for the well-known blood serological data of Taylor, first analyzed by Fisher. This stepwise method provides a technique for eliminating categories that do not contribute significantly to the relationship.     

Enriched biplots for canonical correlation analysis

This paper discusses biplots of the between-set correlation matrix obtained by canonical correlation analysis. It is shown that these biplots can be enriched with the representation of the cases of the original data matrices. A representation of the cases that is optimal in the generalized least squares sense is obtained by the superposition of a scatterplot of the canonical variates on the biplot of the between-set correlation matrix. Goodness of fit statistics for all correlation and data matrices involved in canonical correlation analysis are discussed. It is shown that adequacy and redundancy coefficients are in fact statistics that express the goodness of fit of the original data matrices in the biplot. The within-set correlation matrix that is represented in standard coordinates always has a better goodness of fit than the within-set correlation matrix that is represented in principal coordinates. Given certain scalings, the scalar products between variable vectors approximate correlations better than the cosines of angles between variable vectors. Several data sets are used to illustrate the results.     

Correlated variables in regression: Clustering and sparse estimation

We consider estimation in a high-dimensional linear model with strongly correlated variables. We propose to cluster the variables first and do subsequent sparse estimation such as the Lasso for cluster-representatives or the group Lasso based on the structure from the clusters. Regarding the first step, we present a novel and bottom-up agglomerative clustering algorithm based on canonical correlations, and we show that it finds an optimal solution and is statistically consistent. We also present some theoretical arguments that canonical correlation based clustering leads to a better-posed compatibility constant for the design matrix which ensures identifiability and an oracle inequality for the group Lasso. Furthermore, we discuss circumstances where cluster-representatives and using the Lasso as subsequent estimator leads to improved results for prediction and detection of variables. We complement the theoretical analysis with various empirical results.     

A system for computing joint probabilities from jensen's bivariate f distribution

A system of subroutines is presented for efficient computation of joint probabilities from Jensen's bivariate F distribution. Any valid set of parameters is permitted, whereas previous work was limited to the special case of equal numerator degrees of freedom and equal canonical correlations in the underlying multinormal distribution. The use of joint Probabilities from Jensen's bivariate F distribution is demonstrated via an application to two-way ANOVA without interaction.     

A new space–time multivariate approach for environmental data analysis

Air quality control usually requires a monitoring system of multiple indicators measured at various points in space and time. Hence, the use of space–time multivariate techniques are of fundamental importance in this context, where decisions and actions regarding environmental protection should be supported by studies based on either inter-variables relations and spatial–temporal correlations. This paper describes how canonical correlation analysis can be combined with space–time geostatistical methods for analysing two spatial–temporal correlated aspects, such as air pollution concentrations and meteorological conditions. Hourly averages of three pollutants (nitric oxide, nitrogen dioxide and ozone) and three atmospheric indicators (temperature, humidity and wind speed) taken for two critical months (February and August) at several monitoring stations are considered and space–time variograms for the variables are estimated. Simultaneous relationships between such sample space–time variograms are determined through canonical correlation analysis. The most correlated canonical variates are used for describing synthetically the underlying space–time behaviour of the components of the two sets.     

On a measure of dependence based on fisher's information matrix

A class of measures of dependence between two random vectors is defined, in terms of the canonical correlations obtained from Fisher's information matrix. Some basic properties are proved for this class of measures. Examples are given to illustrate that the class gives good measures, under normal models. Interesting measures are also arise for bivariate models where the correlation coefficient does not exist for some values of the parameters of the model.     

On the robustness of cointegration tests when series are fractionally intergrated

This paper shows that when series are fractionally integrated, but unit root tests wrongly indicate that they are I(1), Johansen likelihood ratio (LR) tests tend to find too much spurious cointegration, while the Engle-Granger test presents a more robust performance. This result holds asymptotically as well as infinite samples. The different performance of these two methods is due to the fact that they are based on different principles. The Johansen procedure is based on maximizing correlations (canonical correlation) while Engle-Granger minimizes variances (in the spirit of principal components).