A CLT for regularized sample covariance matrices
Anderson, Greg W. ; Zeitouni, Ofer
Ann. Statist., Tome 36 (2008) no. 1, p. 2553-2576 / Harvested from Project Euclid
We consider the spectral properties of a class of regularized estimators of (large) empirical covariance matrices corresponding to stationary (but not necessarily Gaussian) sequences, obtained by banding. We prove a law of large numbers (similar to that proved in the Gaussian case by Bickel and Levina), which implies that the spectrum of a banded empirical covariance matrix is an efficient estimator. Our main result is a central limit theorem in the same regime, which to our knowledge is new, even in the Gaussian setup.
Publié le : 2008-12-15
Classification:  Random matrices,  sample covariance,  regularization,  62H12,  15A52
@article{1231165179,
     author = {Anderson, Greg W. and Zeitouni, Ofer},
     title = {A CLT for regularized sample covariance matrices},
     journal = {Ann. Statist.},
     volume = {36},
     number = {1},
     year = {2008},
     pages = { 2553-2576},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1231165179}
}
Anderson, Greg W.; Zeitouni, Ofer. A CLT for regularized sample covariance matrices. Ann. Statist., Tome 36 (2008) no. 1, pp.  2553-2576. http://gdmltest.u-ga.fr/item/1231165179/