Table 4.

Top-1 combinations for the VGG16 in the 15 optimization iterations.

#	Parameters optimizer	Batch size	Dropout ratio	TL learn ratio	Loss	Accuracy	F1	Precision	Recall	AUC	WSM
1	SGD	32	0.57	80%	0.0716	97.21%	97.21%	97.21%	97.21%	0.9966	95.48%
2	SGD	32	0.56	80%	0.0952	96.63%	96.63%	96.63%	96.63%	0.9946	94.74%
3	AdaGrad	32	0.21	30%	0.2933	91.25%	91.25%	91.25%	91.25%	0.9674	89.14%
4	SGD	32	0.58	80%	0.0584	98.08%	98.08%	98.08%	98.08%	0.9967	96.48%
5	AdaGrad	32	0.19	25%	0.1302	95.57%	95.58%	95.57%	95.57%	0.9897	93.57%
6	SGD	32	0.58	80%	0.0394	98.61%	98.61%	98.61%	98.61%	0.9980	97.41%
7	SGD	32	0.58	80%	0.0263	99.13%	99.13%	99.13%	99.13%	0.9991	98.55%
8	SGD	32	0.58	80%	0.0252	99.23%	99.23%	99.23%	99.23%	0.9988	98.73%
9	SGD	32	0.58	80%	0.0221	99.23%	99.23%	99.23%	99.23%	0.9993	99.02%
10	SGD	32	0.58	80%	0.0314	98.99%	98.99%	98.99%	98.99%	0.9995	98.11%
11	SGD	32	0.58	80%	0.0324	98.94%	98.94%	98.94%	98.94%	0.9986	98.01%
12	SGD	32	0.58	80%	0.0226	99.28%	99.28%	99.28%	99.28%	0.9993	99.01%
13	SGD	32	0.58	80%	0.0247	99.09%	99.09%	99.09%	99.09%	0.9988	98.63%
14	SGD	32	0.58	80%	0.0251	99.09%	99.09%	99.09%	99.09%	0.9988	98.60%
15	SGD	32	0.58	80%	0.0292	98.85%	98.85%	98.85%	98.85%	0.9987	98.09%