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

Table 14

Performance of the proposed method, RMSprop with decay.

	ResNet-20 on Cafri-10		LSTM on IMDB
	Val-loss	Val-acc	Val-loss	Val-acc
(RMSprop + d) + SGD	0.6149	0.8464	0.4872	0.8338
(RMSprop + d) + (SGD + d)	0.9866	0.7530	0.5555	0.8289
(RMSprop + d) + (SGD + M)	0.6012	0.8493	0.4848	0.8341
(RMSprop + d) + (SGD + M + d)	0.7868	0.7915	0.5462	0.8257
(RMSprop + d) + RMSprop	0.7120	0.8212	0.7487	0.8161
(RMSprop + d) + (RMSprop + d)	0.9602	0.7771	0.8846	0.7980
(RMSprop + d) + Adam	0.8002	0.7920	0.9164	0.8114
(RMSprop + d) + (Adam + d)	0.8254	0.8000	1.0931	0.8146

The bold values represent the best results.