Table 7 Metric summarize
From: PUMD: a PU learning-based malicious domain detection framework
Metric | Description |
|---|---|
TP | Malicious domains labeled as malicious |
FP | Benign domains labeled as malicious |
TN | Benign domains labeled as benign |
FN | Malicious domains labeled as benign |
ACC | \((TP + TN)/(TP + FP + TN + FN)\) |
Precision | \(TP/(TP + FP)\) |
Recall | \(TP/(TP + FN)\) |
F1 | \(2 \times (precision \cdot recall)/(precision + recall)\) |
TPR | \(TP/(TP + FN)\) |
FPR | \(FP/(FP + TN)\) |
ROC | A curve plotting TPR against FPR with various thresholds |
AUC | Area under the ROC curve |
MCC | \(\frac{(TP \cdot TN - FP \cdot FN)}{\sqrt{(TP + FP)(TP + FN)(TN + FP)(TN + FN)}}\) |