Table 11 Comparison of our method with existing recent relevant works in the literature.

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

Paper

Model

Analysis technique

Feature type

Threat type

Accuracy (%)

21

VGG416

Static

Image

Malware

91.41

21

VGG416

Hybrid

Image

Malware

94.70

12

Xception ColSeq

Static

Image

Malware

98.20

16

ResNet50

Static

Image

Malware

99.50

31

ResNet50

Static

Image

Malware

98.61

32

ResNet50

Static

Image

Malware

98.62

67

Xception CNN

Static

Image

Malware

99.20

8

Logistic Regression (RLR)

Dynamic

CSV

Ransomware

96.34

23

CNN

Dynamic

CSV

Ransomware

95.90

24

Random forest

Dynamic

CSV

Ransomware

97.28

29

CNN, LSTM, MLP

Static + Dynamic

Image + CSV

Ransomware

98.75

28

2-layered CNN

Dynamic

CSV

Ransomware

98.20

Our proposed method

ResNet50

Dynamic

Image (from CSV)

Ransomware

99.96