Table 7 Comparative performance of three pre-trained networks using different ML classifiers

Pretrained model	ML classifier	TP	FP	FN	TN	Precision	Sensitivity	Accuracy	F-score
ResNet-18	SVM	310	46	50	254	0.87079	0.86111	0.85455	0.86592
	DT	294	72	66	228	0.80328	0.81667	0.79091	0.80992
	KNN	308	56	52	244	0.84615	0.85556	0.83636	0.85083
	Naïve bayes	281	42	79	258	0.86997	0.78056	0.81667	0.82284
VGG-19	SVM	297	64	63	236	0.82271	0.8250	0.80758	0.82386
	DT	261	79	99	221	0.76765	0.7250	0.7303	0.74571
	KNN	292	81	68	219	0.78284	0.81111	0.77424	0.79673
	Naïve bayes	257	79	103	221	0.76488	0.71389	0.72424	0.73851
MobileNet V2	SVM	334	23	26	277	0.93557	0.92778	0.92575	0.93165
	DT	282	77	78	223	0.78552	0.78333	0.76515	0.78442
	KNN	320	66	40	234	0.82902	0.88889	0.83939	0.83939
	Naïve bayes	225	73	135	224	0.75503	0.625	0.75667	0.68389

Quick links

Search