Skip to main content

Table 12 Results of evaluation

From: Big knowledge-based semantic correlation for detecting slow and low-level advanced persistent threats

Actual class\Predicted class

APT

Benign

Total

Recognition(%)

APT

8(TP)

1 (FN)

9(P)

88.88 (Sensitivity= \(\frac{TP}{P}\))

Benign

12(FP)

98(TN)

110(N)

89.09 (Specificity= \(\frac{TN}{N}\))

Total

20

99

119

89.07(Accuracy= \(\frac{TP+TN}{P+N}\))

Other metrics(%)

40.00 (Precision= \(\frac{TP}{TP+FP}\))

88.88 (Recall= \(\frac{TP}{TP+FN}\))

55.17 (F-Measure)

10.92 (Error-rate= \(\frac{FP+FN}{P+N}\))