Software Defect Prediction for Healthcare Big Data: An Empirical Evaluation of Machine Learning Techniques

Table 7

Comparative analysis of TPR and FPR of ML technique on different datasets.

Dataset		SVM	J48	RF	MLP	RBF	HMM	CDT	A1DE	NB	KNN
AR1	TPR	0.917	0.901	0.901	0.901	0.917	0.926	0.926	0.909	0.851	0.901
	FPR	0.926	0.723	0.928	0.723	0.926	0.926	0.926	0.927	0.523	0.621
AR3	TPR	0.889	0.873	0.921	0.937	0.873	0.873	0.873	0.921	0.905	0.857
	FPR	0.657	0.446	0.332	0.33	0.766	0.873	0.873	0.332	0.227	0.555
CM1	TPR	0.896	0.88	0.892	0.876	0.896	0.902	0.894	0.863	0.853	0.847
	FPR	0.902	0.849	0.848	0.886	0.902	0.902	0.902	0.869	0.616	0.762
JM1	TPR	0.817	0.799	0.827	0.82	0.82	0.183	0.817	0.815	0.814	0.771
	FPR	0.812	0.631	0.635	0.77	0.757	0.183	0.695	0.662	0.658	0.551
KC2	TPR	0.828	0.814	0.833	0.847	0.837	0.795	0.83	0.833	0.835	0.805
	FPR	0.634	0.422	0.431	0.435	0.472	0.795	0.439	0.424	0.473	0.432
KC3	TPR	0.82	0.794	0.814	0.773	0.799	0.186	0.82	0.789	0.789	0.722
	FPR	0.792	0.562	0.707	0.609	0.797	0.186	0.663	0.561	0.52	0.728
MC1	TPR	0.993	0.994	0.995	0.994	0.993	0.993	0.994	0.982	0.942	0.995
	FPR	0.993	0.701	0.657	0.73	0.993	0.993	0.774	0.628	0.38	0.496