WeSpeR: Computing non-linear shrinkage formulas for the weighted sample covariance
Statistics Theory
2025-09-03 v2 Machine Learning
Probability
Computation
Machine Learning
Statistics Theory
Abstract
We address the issue of computing the non-linear shrinkage formulas for the weighted sample covariance in high dimension. We use theoretical properties of the asymptotic sample spectrum in order to derive the \textit{WeSpeR} algorithm and significantly speed up non-linear shrinkage in dimension higher than . Empirical tests confirm the good properties of the \textit{WeSpeR} algorithm. We provide the implementation in PyTorch for it.
Cite
@article{arxiv.2410.14413,
title = {WeSpeR: Computing non-linear shrinkage formulas for the weighted sample covariance},
author = {Benoit Oriol},
journal= {arXiv preprint arXiv:2410.14413},
year = {2025}
}