Table 5.

Experimental results on the Skin-lesion dataset.

Method	TPR	TNR	PPV	NPV	ACC	${\mathit{F}}_{\mathit{l}}^{\mathit{p}}$	${\mathit{F}}_{\mathit{l}}^{\mathit{n}}$	MCC
Texture analysis [17]	0.87	0.71	0.76	-	0.75	-	-	-
HLIFs [16]	0.96	0.73	-	-	0.83	-	-	-
BIBS [15]	0.92	0.88	0.91	-	0.90	-	-	-
Decision Support [14]	0.84	0.79	-	-	0.81	-	-	-
Color pigment boundary [13]	0.95	0.88	0.92	-	0.82	-	-	-
R. Amelard et al. [43] Asymmetry $F_C$	0.73	0.64	-	-	0.69	-	-	-
R. Amelard et al. [43] Proposed HLIFs	0.79	0.68	-	-	0.75	-	-	-
R. Amelard et al. [43] Cavalcanti feature set	0.84	0.78	-	-	0.82	-	-	-
R. Amelard et al. [43] Modified $F_C$	0.86	0.75	-	-	0.72	-	-	-
R. Amelard et al. [43] Combined $F_{MC}$ $F_A^{HLIFS}$	0.91	0.80	-	-	0.86	-	-	-
Resnet50+Resnet18	0.84	0.92	0.91	0.85	0.88	0.87	0.88	0.76
Resnet50+Googlenet+Alexnet	0.87	0.65	0.71	0.84	0.76	0.78	0.73	0.54