Skip to main content

Table 8 Performance of different feature subsets

From: IGRF-RFE: a hybrid feature selection method for MLP-based network intrusion detection on UNSW-NB15 dataset

Subsets

Num.

Precision (%)

Recall (%)

F1 score (%)

Accuracy (%)

All Features

42

80.37

82.25

80.22

82.25

IG

25

82.24

83.13

81.49

83.13

RF

22

82.43

83.42

81.62

83.42

IG &RF Uni.

27

83.30

80.60

80.36

80.60

IG &RF Inter.

20

81.84

82.90

81.67

82.90

IGRF-RFE

23

83.60

84.24

82.85

84.24