A Matrix Variate Generalized Hyperbolic Distribution

Three way data can be modelled using a matrix variate distribution. Although ample work has been done in the literature regarding the matrix variate normal distribution, there is relative paucity for skewed distributions. Herein, we present a formulation of the matrix variate generalized hyperbolic distribution based on a matrix normal variance-mean mixture model. Parameter estimation is discussed, and simulated data are used for illustration.