Table 11.

Top-1 combinations for the MobileNetV2 in the 15 optimization iterations.

#	Parameters optimizer	Batch size	Dropout ratio	TL learn ratio	Loss	Accuracy	F1	Precision	Recall	AUC	WSM
1	AdaDelta	32	0.28	0%	0.7376	64.21%	64.22%	64.21%	64.21%	0.6683	62.68%
2	AdaDelta	32	0.28	0%	0.7386	64.41%	64.41%	64.41%	64.41%	0.6675	62.87%
3	AdaDelta	32	0.28	0%	0.7368	64.55%	64.56%	64.55%	64.55%	0.6690	63.01%
4	AdaDelta	32	0.28	0%	0.7388	64.17%	64.17%	64.17%	64.17%	0.6663	62.63%
5	AdaDelta	32	0.28	0%	0.7367	64.60%	64.60%	64.60%	64.60%	0.6687	63.05%
6	AdaDelta	32	0.28	0%	0.7360	64.55%	64.56%	64.55%	64.55%	0.6692	63.01%
7	AdaDelta	32	0.28	0%	0.7384	64.69%	64.70%	64.69%	64.69%	0.6684	63.15%
8	AdaDelta	32	0.28	0%	0.7372	64.74%	64.75%	64.74%	64.74%	0.6690	63.19%
9	AdaDelta	32	0.28	0%	0.7391	64.74%	64.75%	64.74%	64.74%	0.6678	63.19%
10	AdaDelta	32	0.28	0%	0.7369	64.55%	64.56%	64.55%	64.55%	0.6685	63.01%
11	AdaDelta	32	0.28	0%	0.7384	64.55%	64.56%	64.55%	64.55%	0.6679	63.01%
12	AdaDelta	32	0.28	0%	0.7378	64.55%	64.56%	64.55%	64.55%	0.6681	63.01%
13	AdaDelta	32	0.28	0%	0.7374	64.74%	64.75%	64.74%	64.74%	0.6687	63.19%
14	AdaDelta	32	0.28	0%	0.7372	64.60%	64.60%	64.60%	64.60%	0.6688	63.05%
15	AdaDelta	32	0.28	0%	0.7395	64.50%	64.51%	64.50%	64.50%	0.6675	62.96%