An Informative and Comprehensive Behavioral Characteristics Analysis Methodology of Android Application for Data Security in Brain-Machine Interfacing

Table 5

Performance of dynamic behavioral characteristic combination.

Behavioral characteristic set	Algorithm	TP	FP	Precision	Recall	F-Measure	ROC area
Quantitative	SVM	0.914	0.086	0.914	0.914	0.914	0.905
	KNN	0.909	0.091	0.908	0.928	0.908	0.889
	Random forest	0.927	0.073	0.927	0.927	0.927	0.916
Timing	SVM	0.875	0.125	0.858	0.875	0.866	0.5
	KNN	0.875	0.125	0.858	0.875	0.866	0.5
	Random forest	0.881	0.119	0.881	0.881	0.881	0.59
Semantic	SVM	0.91	0.09	0.91	0.91	0.91	0.899
	KNN	0.902	0.092	0.902	0.902	0.902	0.881
	Random forest	0.917	0.093	0.917	0.917	0.917	0.903
System call	SVM	0.855	0.145	0.856	0.856	0.856	0.879
	KNN	0.812	0.188	0.815	0.815	0.815	0.846
	Random forest	0.871	0.129	0.871	0.87	0.871	0.892
All	SVM	0.945	0.055	0.945	0.938	0.948	0.966
	KNN	0.947	0.053	0.947	0.942	0.945	0.961
	Random forest	0.965	0.035	0.979	0.977	0.979	0.986