Table 6 Comparison of experimental results(Microsoft Malware Classification Challenge).

From: GCSA-ResNet: a deep neural network architecture for Malware detection

Model

accuracy

precision

Recall-mic

Recall-mac

F1 -mic

F1-mac

F1-wei

SPC

FPR

NPV

DE

Time

CNN

90.86

90.86

90.53

85.00

90.86

87.12

90.92

98.83

1.17

98.84

97.97

1197

CNN-SVM

93.92

93.92

93.92

91.49

93.92

91.65

93.95

99.22

0.78

99.22

98.65

1263

VIT

88.48

88.48

88.48

86.13

88.48

85.08

88.51

98.52

1.48

98.52

97.44

3162

Resnet50

97.53

97.53

97.53

95.91

97.53

95.97

97.53

99.68

0.32

99.68

99.45

1802

SE-Resnet

98.26

98.26

98.26

98.16

98.26

98.10

98.27

99.78

0.22

99.77

99.61

2185

CBAM-Resnet

97.80

97.81

97.81

97.68

97.81

97.63

97.80

99.72

0.28

99.72

99.51

2241

GCSA-Resnet

98.58

98.58

98.58

98.35

98.58

98.37

98.58

99.82

0.18

99.82

99.69

2413