Table 2.

Comparison of the six AI models on the basis of multiple classification metrics

Arch*	Sens	Spec	Prec	NPR	FPR	FDR	FNR	F1	MCC	Kappa
k-NN	0.5097	0.9099	0.798	0.7266	0.0901	0.2020	0.4903	0.6220	0.4692	0.444
RF	0.9065	0.9926	0.9798	0.964	0.0074	0.0202	0.0935	0.9417	0.9212	0.920
IV3	0.8624	0.9813	0.9495	0.946	0.0187	0.0505	0.1376	0.9038	0.8692	0.867
VGG19	0.9899	0.9964	0.9899	0.9964	0.0036	0.0101	0.0101	0.9899	0.9863	0.986
CNN	0.9899	0.9964	0.9899	0.9964	0.0036	0.0101	0.0101	0.9899	0.9863	0.986
iCNN	0.9899	0.9964	0.9899	0.9964	0.0036	0.0101	0.0101	0.9899	0.9863	0.986

*Arch: architecture; Sens: sensitivity; Spec: specificity; Prec: precision MCC: Mathew’s correlation coefficient; F1: F1-score; IV3: InceptionV3;