Optimal cleaning for singular values of cross-covariance matrices

01/16/2019
by   Florent Benaych-Georges, et al.
0

We give a new algorithm for the estimation of the cross-covariance matrix E XY' of two large dimensional signals X∈R^n, Y∈R^p in the context where the number T of observations of the pair (X,Y) is itself large, but with n,p non negligible with respect to T. This algorithm is optimal among rotationally invariant estimators, i.e. estimators derived from the empirical estimator by cleaning the singular values, while letting singular vectors unchanged. We give an interpretation of the singular value cleaning in terms of overfitting ratios.

READ FULL TEXT
research
12/18/2019

Cross product-free matrix pencils for computing generalized singular values

It is well known that the generalized (or quotient) singular values of a...
research
03/07/2021

Cascaded Filtering Using the Sigma Point Transformation (Extended Version)

It is often convenient to separate a state estimation task into smaller ...
research
08/01/2020

Simpler Proofs for Approximate Factor Models of Large Dimensions

Estimates of the approximate factor model are increasingly used in empir...
research
12/19/2019

Singular matrix variate Birnbaum-Saunders distribution under elliptical models

This work sets the matrix variate Birnbaum-Saunders theory in the contex...
research
05/15/2023

Matrix quadratic risk of orthogonally invariant estimators for a normal mean matrix

In estimation of a normal mean matrix under the matrix quadratic loss, w...
research
10/10/2018

Random matrix-improved estimation of covariance matrix distances

Given two sets x_1^(1),...,x_n_1^(1) and x_1^(2),...,x_n_2^(2)∈R^p (or C...
research
09/25/2020

ScreeNOT: Exact MSE-Optimal Singular Value Thresholding in Correlated Noise

We derive a formula for optimal hard thresholding of the singular value ...

Please sign up or login with your details

Forgot password? Click here to reset