17 October 2014 in Biometrika
Robust estimators of the restricted covariance matrices associated with elliptical graphical models are studied. General asymptotic results, which apply to both decomposable and nondecomposable graphical models, are presented for robust plug-in type estimators. These extend results previously established only for the decomposable case. Furthermore, a class of graphical M-estimators for the restricted covariance matrices is introduced and compared with the corresponding plug-in M-estimators. The two approaches are shown to be asymptotically equivalent under random sampling from an elliptical distribution. A simulation study demonstrates the superiority of the graphical M-estimators for small samples.
21 February 2013 in arXiv
Asymptotic properties of scatter estimators for elliptical graphical models are studied. Such models impose a given pattern of zeros on the inverse of the shape matrix of an elliptically distributed random vector. In particular, we introduce the class of graphical M-estimators and compare them to plug-in M-estimators. It turns out that, under suitable conditions, both approaches yield the same asymptotic efficiency. Furthermore, the results of this paper apply to both decomposable and non-decomposable graphical models and so generalize the results for decomposable models given by Vogel & Fried (2011) for the plug-in M-estimators.