loss Module¶
This module provides loss functions frequently encountered in the literature on high dimensional covariance matrix estimation.
Functions¶
|
Squared Frobenius norm scaled by 1/p. |
|
The minimum variance loss function of Ledoit and Wolf (2018). |
|
The Marchenko-Pastur distribution. |
|
The percentage relative improvement in average loss (PRIAL) over the sample covariance matrix. |