Some Simplifying Results on BLUEs
James M. Lowerre · Journal of the American Statistical Association · 1977
When the error covariance matrix, V, for the model , is nonsingular it is known that the Best Linear Unbiased Estimator (BLUE) of is , in which M+ represents the Moore-Penrose generalized inverse. When V is singular this does not generalize to . This paper discusses approaches to finding blues. Simple proofs and new results are presented, primarily based on a study of the equation blue , where w′Y is a proper random variable and a is a scalar constant.