[Retracted] Gradient Descent Optimization in Deep Learning Model Training Based on Multistage and Method Combination Strategy

Table 11

Performance of the proposed method, SGD with Momentum.

	ResNet-20 on Cafri-10		LSTM on IMDB
	Val-loss	Val-acc	Val-loss	Val-acc
(SGD + M) + SGD	0.6313	0.8359	0.3745	0.8170
(SGD + M) + (SGD + M)	0.6970	0.8191	0.4320	0.8130
(SGD + M) + (SGD + d)	0.6468	0.8321	0.3924	0.8225
(SGD + M) + (SGD + M + d)	0.7518	0.7969	0.3808	0.8318
(SGD + M) + RMSprop	0.8214	0.7911	0.4959	0.8300
(SGD + M) + (RMSprop + d)	1.0059	0.7338	0.4626	0.8325
(SGD + M) + Adam	0.9730	0.7499	0.8488	0.8143
(SGD + M) + (Adam + d)	0.8190	0.7899	0.9352	0.8083

The bold values represent the best results.