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

Table 10

Performance of the proposed method, Adam.

	ResNet-20 on Cafri-10		LSTM on IMDB
	Val-loss	Val-acc	Val-loss	Val-acc
Adam + SGD	0.6088	0.8494	0.9167	0.8135
Adam + (SGD + M)	0.6582	0.8335	1.0421	0.8156
Adam + (SGD + d)	0.6108	0.8451	0.9032	0.8140
Adam + (SGD + M + d)	0.7453	0.8093	1.1045	0.8150
Adam + RMSprop	0.6929	0.8304	1.1457	0.8089
Adam + (RMSprop + d)	0.8948	0.7816	1.1166	0.8038
Adam + Adam	0.8138	0.7999	1.2044	0.8060
Adam + (Adam + d)	1.1411	0.7164	1.3086	0.8089

The bold values represent the best results.