A Novel Malware Detection and Family Classification Scheme for IoT Based on DEAM and DenseNet

Table 10

Precision, recall, and F1 score of DenseNet on BIG 2015 dataset.

	Ramnit	Lollipop	Kelihos_ver3	Vundo	Simda	Tracur	Kelihos_ver1	Obfuscator.ACY	Gatak	Precision	Recall	F1 score
Ramnit	285	8	0	0	1	1	3	11	0	0..976	0.922	0.948
Lollipop	0	495	0	1	0	0	0	0	0	0.948	0.997	0.972
Kelihos_ver3	0	0	588	0	0	0	0	0	0	0.989	1.000	0.994
Vundo	0	2	0	90	0	1	0	1	1	0.957	0.947	0.952
Simda	0	0	0	1	5	0	1	0	1	0.833	0.625	0.714
Tracur	3	2	2	1	0	135	2	0	5	0.978	0.900	0.937
Kelihos_ver1	0	1	2	0	0	0	76	0	0	0.915	0.962	0.938
Obfuscator.ACY	4	8	2	0	0	1	1	224	5	0.941	0.914	0.927
Gatak	0	6	0	1	0	0	0	2	193	0.941	0.955	0.948
Macro										0.942	0.914	0.926