In the general linear model $\mathscr{E}(\mathbf{y}) = \mathbf{X\beta}$, the vector $\mathbf{A\beta}$ is estimable whenever there is a matrix $\mathbf{B}$ so that $\mathscr{E}(\mathbf{By}) = \mathbf{A\beta}$. Several characterizations of estimability are presented along with short easy proofs. The characterizations involve rank equalities, generalized inverses, Schur complements and partitioned matrices.
@article{1176344564,
author = {Alalouf, I. S. and Styan, G. P. H.},
title = {Characterizations of Estimability in the General Linear Model},
journal = {Ann. Statist.},
volume = {7},
number = {1},
year = {1979},
pages = { 194-200},
language = {en},
url = {http://dml.mathdoc.fr/item/1176344564}
}
Alalouf, I. S.; Styan, G. P. H. Characterizations of Estimability in the General Linear Model. Ann. Statist., Tome 7 (1979) no. 1, pp. 194-200. http://gdmltest.u-ga.fr/item/1176344564/