Table 4.

Performance metrics obtained from the confusion matrix.

Metric	Equation	Short Description
TPR (DR)	$\frac{\mathrm{TP}}{\mathrm{TP}+\mathrm{FN}}$	The performance of the CNN model in detecting malicious samples.
Specificity	$\frac{\mathrm{TN}}{\mathrm{TN}+\mathrm{FP}}$	The performance of the CNN model in detecting honest samples.
Precision	$\frac{\mathrm{TP}}{\mathrm{TP}+\mathrm{FP}}$	The ratio of malicious samples predicted as malicious to all malicious samples.
F1-score	$\frac{2\mathrm{TP}}{2\mathrm{TP}+\mathrm{FP}+\mathrm{FN}}$	Expressed as the harmonic mean between precision and sensitivity.
Accuracy	$\frac{\mathrm{TP}+\mathrm{TN}}{\mathrm{TP}+\mathrm{FP}+\mathrm{FN}+\mathrm{TN}}$	Measures the overall correctness of predictions by the model.
FPR	$\frac{\mathrm{TP}}{\mathrm{FP}+\mathrm{TN}}$	Measures the proportion of actual honest that were incorrectly classified as malicious samples.