Table 8.

Prediction results of the benchmark models

Model	Epoch	Batch size	Label	Precision	Recall	F1-Score	Accuracy	Confusion matrix
EfficientNet B0	3	32	COVID-19	0.98	0.97	0.98	84.8%	249	5	2	0
			Normal	0.92	0.94	0.94		1	632	13	8
			Pneumonia Bacterial	0.76	0.88	0.82		2	26	530	42
			Pneumonia Viral	0.75	0.46	0.57		1	24	155	151
EfficientNet B2	5	32	COVID-19	0.99	0.98	0.99	85.7%	251	3	1	1
			Normal	0.94	0.96	0.95		0	630	8	16
			Pneumonia Bacterial	0.78	0.86	0.82		1	17	516	66
			Pneumonia Viral	0.69	0.55	0.61		0	17	133	181
InceptionV3	5	64	COVID-19	0.98	0.96	0.97	85.0%	245	6	5	0
			Normal	0.92	0.97	0.94		1	633	5	15
			Pneumonia Bacterial	0.79	0.84	0.81		3	26	503	68
			Pneumonia Viral	0.69	0.55	0.61		0	24	124	183
InceptionResNetV2	5	32	COVID-19	0.98	0.96	0.97	83.5%	245	1	8	2
			Normal	0.94	0.95	0.95		2	622	8	22
			Pneumonia Bacterial	0.74	0.89	0.81		3	18	532	47
			Pneumonia Viral	0.66	0.42	0.51		0	20	173	138
MobileNetV2	5	64	COVID-19	0.99	0.96	0.97	84.6%	245	6	1	4
			Normal	0.93	0.98	0.95		0	642	0	12
			Pneumonia Bacterial	0.80	0.81	0.80		1	20	484	95
			Pneumonia Viral	0.63	0.56	0.59		1	25	118	187
NASNetMobile	3	32	COVID-19	0.96	0.96	0.96	84.0%	245	5	6	0
			Normal	0.91	0.95	0.93		6	620	15	13
			Pneumonia Bacterial	0.76	0.87	0.81		1	26	522	51
			Pneumonia Viral	0.71	0.48	0.57		2	28	142	159
ResNetV2_152	5	64	COVID-19	0.98	0.94	0.96	85.5%	241	7	5	3
			Normal	0.93	0.98	0.95		2	638	2	12
			Pneumonia Bacterial	0.81	0.81	0.81		3	22	487	88
			Pneumonia Viral	0.67	0.63	0.65		1	16	106	208
VGG16	3	32	COVID-19	0.93	0.83	0.88	79.4%	212	9	34	1
			Normal	0.88	0.92	0.90		9	604	23	18
			Pneumonia Bacterial	0.71	0.84	0.77		3	40	503	54
			Pneumonia Viral	0.66	0.43	0.52		4	36	149	142
VGG19	5	64	COVID-19	0.95	0.91	0.93	79.5%	233	14	9	0
			Normal	0.82	0.97	0.89		3	637	10	4
			Pneumonia Bacterial	0.72	0.84	0.77		3	69	505	23
			Pneumonia Viral	0.77	0.27	0.40		5	55	182	89

The best values are shown in bold