Oberseminar Stochastik und Data Science : Martin Metodiev

Title: Estimating large covariance matrices

Abstract: We construct an estimator for high-dimensional covariance matrices when the number of available data points is low with respect to the number of variables. We only require the estimation of a small number of parameters, which are easy to interpret and which can be selected using standard model-choice procedures such as the BIC and the inclusion of interaction effects. The estimator is consistent under no assumptions on the covariance structure, and asymptotically normal with available confidence regions.

https://www.math.fau.de/stochastik/oberseminar-stochastik/