“…In that case, Σ −1 , the inverse of the covariance matrix Σ, is often called the precision matrix. This classical multivariate setting has been studied by Efron and Morris [4], Haff [7], Dey [2], Krishnamoorthy and Gupta [13], Dey et al [3], Zhou et al [19], and Tsukuma and Konno [17]. Note that, in these papers, S is assumed to be invertible.…”