Table 6 Cyberattack detection of SAMFL-SCDCOA model on the UNSW-NB15 dataset.

From: Leveraging self attention driven gated recurrent unit with crocodile optimization algorithm for cyberattack detection using federated learning framework

Class labels

\(Acc{u}_{y}\)

\(Pre{c}_{n}\)

\(Rec{a}_{l}\)

\({F}_{score}\)

\(AU{C}_{score}\)

TRPH (70%)

Normal

98.99

97.02

97.79

97.40

98.53

Generic

98.91

96.93

97.53

97.23

98.39

Exploits

98.81

96.30

97.52

96.91

98.32

Fuzzers

98.88

96.35

97.80

97.07

98.47

Backdoors

99.18

97.11

96.71

96.91

98.13

Shellcode

98.82

94.32

92.07

93.18

95.77

Worms

99.37

90.48

41.76

57.14

70.86

Average

99.00

95.50

88.74

90.83

94.07

TSPH (30%)

Normal

99.00

96.78

97.96

97.37

98.60

Generic

99.15

98.18

97.22

97.70

98.41

Exploits

98.80

96.03

97.91

96.96

98.46

Fuzzers

98.95

96.81

97.93

97.37

98.57

Backdoors

99.08

96.08

97.17

96.62

98.27

Shellcode

98.90

95.68

91.45

93.51

95.53

Worms

99.41

86.36

48.72

62.30

74.32

Average

99.04

95.13

89.77

91.69

94.59