Sun, Xiaoqian; Zhou, Xian - In: Statistics & Probability Letters 78 (2008) 2, pp. 127-134
Suppose that n independent observations are drawn from a multivariate normal distribution Np([mu],[Sigma]) with both mean vector [mu] and covariance matrix [Sigma] unknown. We consider the problem of estimating the precision matrix [Sigma]-1 under the squared loss . It is well known that the...