Skip to contents

Calculate various measures to evaluate the performance of the estimated precision matrix.

Usage

performance(Omega, Sigma = NULL, hatOmega)

Arguments

Omega

The precision matrix.

Sigma

The covariance matrix (default = NULL).

hatOmega

The estimated precision matrix.

Value

A list containing the following components:

Fnorm1

Frobenius (Hilbert-Schmidt) norm between the true and estimated precision matrices.

Fnorm2

Frobenius (Hilbert-Schmidt) norm between the product of the true covariance matrix and estimated precision matrix and the identity matrix.

KL

Kullback-Leibler divergence between the true and estimated precision matrices.

Ql

Quadratic loss between the diagonal elements of the product of the true covariance matrix and estimated precision matrix.

Snorm

Spectral (operator) norm of the difference between the true and estimated precision matrices.

Precision

Precision measure, the ratio of true positives to the total predicted positives.

Recall

Recall measure, also known as Sensitivity, the ratio of true positives to the total actual positives.

Specificity

Specificity measure, the ratio of true negatives to the total actual negatives.

F1

F1 score, the harmonic mean of Precision and Recall.

MCC

Matthews correlation coefficient, a measure of the quality of binary classifications.

Sparsity

The ratio of zero entries to total entries in the estimated precision matrix.