Skip to main content

Table 2 Performance measures used

From: Feature combination networks for the interpretation of statistical machine learning models: application to Ames mutagenicity

Name

Equation

Usage

Balanced accuracy (BAC)

SEN + SPEC 2

Accuracy measure that accounts for bias in the data.

Sensitivity (SEN)

TP TP + FN

Model’s ability to correctly identify positives

Specificity (SPEC)

TN TN + FP

Model’s ability to correctly identify negatives

Coverage (COV)

in domain structures Number of structures * 100

Model’s coverage of the given validation set representing the applicability of the model.