A Malware Detection Scheme Based on Mining Format Information

Table 3

Experimental results for all classification algorithms (feature selection from the whole dataset).

Feature selection method	The number of features	Algorithm	TPR (%)	FPR (%)	Accuracy (%)	AUC
Filter	19	J48	98.9	1.4	98.7	0.994
		Random forest	99.1	1.4	98.9	0.996
		Adaboost (J48)	99.0	1.0	99.0	0.998
		Bagging (J48)	98.9	1.3	98.8	0.997
Wrapper	20	Random forest	99.1	1.0	99.1	0.998