Estimation of large covariance matrices via free deconvolution: computational and statistical aspects
From MaRDI portal
Publication:6435969
arXiv2305.05646MaRDI QIDQ6435969FDOQ6435969
Authors: Reda Chhaibi, Fabrice Gamboa, Mohamed Slim Kammoun, Mauricio Velasco 
Publication date: 9 May 2023
Abstract: The estimation of large covariance matrices has a high dimensional bias. Correcting for this bias can be reformulated via the tool of Free Probability Theory as a free deconvolution. The goal of this work is a computational and statistical resolution of this problem. Our approach is based on complex-analytic methods methods to invert -transforms. In particular, one needs a theoretical understanding of the Riemann surfaces where multivalued transforms live and an efficient computational scheme.
This page was built for publication: Estimation of large covariance matrices via free deconvolution: computational and statistical aspects
Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q6435969)
