Table 9 Accuracy and loss or error rate comparisons between color and Gray image datasets.

From: A novel technique for ransomware detection using image based dynamic features and transfer learning to address dataset limitations

Model

Accuracy (Color image)

Accuracy (Gray image)

Difference (%)

Loss (Color dataset)

Loss (Grayscale dataset)

Difference (%)

ResNet50

99.96

99.91

0.05

0.0026

0.0059

0.0033

EfficientNetB0

99.78

99.70

0.08

0.0069

0.0167

0.0098

Xception

98.91

98.12

0.79

0.0253

0.0471

0.0218

InceptionV3

98.84

98.04

0.80

0.0579

0.1261

0.0682

VGG16

97.84

97.60

0.24

0.0602

0.1179

0.0577

VGG19

97.47

97.40

0.07

0.0874

0.1194

0.0320

CNN-GRU

97.49

97.07

0.42

0.0579

0.0680

0.0101

CNN-LSTM

96.94

96.80

0.14

0.0963

0.0807

−0.0156

CNN

96.84

96.71

0.13

0.0998

0.1045

0.0047

ANN

90.48

89.24

1.24

0.2707

0.2853

0.0146

Back to article page