Research Article

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

Table 13

Performance of the proposed method, SGD with Momentum and decay.

ResNet-20 on Cafri-10LSTM on IMDB
Val-lossVal-accVal-lossVal-acc

(SGD + M + d) + SGD0.63240.83400.37360.8366
(SGD + M + d) + (SGD + M)0.85330.77370.37700.8369
(SGD + M + d) + (SGD + d)0.64850.82780.39700.8234
(SGD + M + d) + (SGD + M + d)0.82890.78680.42740.8036
(SGD + M + d) + RMSprop0.97950.75930.49490.8245
(SGD + M + d) + (RMSprop + d)0.83710.78840.45870.8279
(SGD + M + d) + Adam1.27220.70580.85690.8091
(SGD + M + d) + (Adam + d)0.74810.81120.90560.8070

The bold values represent the best results.