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

Table 1

The categories of common optimization methods.

	Learning rate adjustment method	Gradient estimation optimization method
Momentum		✓
Nesterov Momentum		✓
Adagrad		✓
Adadelta		✓
RMSprop		✓
Adam		✓
ReduceLROnPlateau	✓
Piecewise constant	✓
Exponential decay	✓
Natural exponential decay	✓
Polynomial decay	✓
Cosine decay	✓
Gradual warmup	✓