On the cumulants of affine equivariant estimators in elliptical families (Q2277712)

The authors consider a statistical model for data which take values in \(R^ d\) and having elliptically distributed errors. They obtain expressions for covariances of affine equivariant estimators \({\hat \mu}\) and \({\hat \Sigma}\) of the mean vector \(\mu\) and the scatter matrix \(\Sigma\) in terms of their expectations and some unknown constants that depend on the model and the estimator. Higher order cumulants are also developed. These expressions are obtained using tensor methods. Inter alia, the authors develop the tensor algebra machinery required to derive the main result that characterizes orthogonal invariant tensors. Application of the basic result to multivariate location and scatter and multivariate regression are discussed.