Nonlinear Shrinkage: An Introduction

High Dimensional Statistics

Optimal covariance matrix estimation in high dimensions

Many applications in statistics, machine learning, and other areas such as finance and biology, need an accurate estimate of a covariance matrix. However, nowadays many of these applications involve high dimensional data and so the usual (sample) covariance estimator just doesn’t cut it anymore. Thus a myriad of papers over the last two decades have been trying to tackle exactly this problem. An extremely powerful method that has emerged from …

covariance-matrix editors pick high-dimensional-data introduction machine learning shrinkage