Likelihood-Based Inference with Separable Correlation Matrices

Karl Oskar Ekvall

This paper proposes methods for likelihood-based inference in multivariate linear regressions when the correlation matrix of the responses is separable; that is, it has a Kronecker…