Binary File’s Visualization and Entropy Features Analysis Combined with Multiple Deep Learning Networks for Malware Classification

Table 9

Proposed method compared with other malware classification methods on the Big2015 dataset.

Methods	Acc (%)	Recall (%)	(%)	RT (s)
GIST + KNN [3]	94.82	94.81	94.75	0.621
LBP + KNN [25]	91.87	91.88	91.87	0.045
GIST + DSIFT + RF [6]	95.09	95.10	95.06	0.004
VGG16 (fine-tune) [28]	97.92	97.93	97.92	62.59
Our method	99.54	99.53	99.53	7.91

RT means running time.