Explainable Artificial Intelligence-Based IoT Device Malware Detection Mechanism Using Image Visualization and Fine-Tuned CNN-Based Transfer Learning Model

Table 4

Classwise performance.

Dataset 2
Families	Precision	Recall	F1-score
Addisplay	0.91	0.89	0.9
Addisplay + + adware	0.89	0.91	0.9
Adload	0.92	0.92	0.92
Adsware	0.9	0.9	0.9
Adware + + adware	0.9	0.91	0.91
Adware + + grayware + + virus	0.88	0.88	0.88
Adware + + virus	0.91	0.9	0.9
Adwareare	0.92	0.92	0.92
Backdoor	0.92	0.92	0.92
Banker + + trojan	0.94	0.91	0.92
Click	0.92	0.92	0.92
Clicker	0.88	0.91	0.9
Clicker + + trojan	0.89	0.87	0.88
Clickfraud + + riskware	0.91	0.91	0.91
Exploit	0.86	0.87	0.86
Fakeangry	0.88	0.9	0.89
Fakeapp	0.94	0.9	0.92
Fakeapp + + trojan	0.96	0.95	0.95
Fakeinst + + trojan	0.96	0.96	0.96
Average	0.91	0.91	0.91
Dataset 1
Benign	0.99	0.99	0.99
Malware	0.99	0.99	0.99
Average	0.99	0.99	0.99