Covariance regularization by thresholding
Bickel, Peter J. ; Levina, Elizaveta
Ann. Statist., Tome 36 (2008) no. 1, p. 2577-2604 / Harvested from Project Euclid
This paper considers regularizing a covariance matrix of p variables estimated from n observations, by hard thresholding. We show that the thresholded estimate is consistent in the operator norm as long as the true covariance matrix is sparse in a suitable sense, the variables are Gaussian or sub-Gaussian, and (log p)/n→0, and obtain explicit rates. The results are uniform over families of covariance matrices which satisfy a fairly natural notion of sparsity. We discuss an intuitive resampling scheme for threshold selection and prove a general cross-validation result that justifies this approach. We also compare thresholding to other covariance estimators in simulations and on an example from climate data.
Publié le : 2008-12-15
Classification:  Covariance estimation,  regularization,  sparsity,  thresholding,  large p small n,  high dimension low sample size,  62H12,  62F12,  62G09
@article{1231165180,
     author = {Bickel, Peter J. and Levina, Elizaveta},
     title = {Covariance regularization by thresholding},
     journal = {Ann. Statist.},
     volume = {36},
     number = {1},
     year = {2008},
     pages = { 2577-2604},
     language = {en},
     url = {http://dml.mathdoc.fr/item/1231165180}
}
Bickel, Peter J.; Levina, Elizaveta. Covariance regularization by thresholding. Ann. Statist., Tome 36 (2008) no. 1, pp.  2577-2604. http://gdmltest.u-ga.fr/item/1231165180/