Quantitative Prediction of Low-Permeability Sandstone Grain Size Based on Conventional Logging Data by Deep Neural Network-Based BP Algorithm

Table 3

List of optimization parameters for neural network training.

Hyperparameter	Model 1	Model 2	Model 3	Model 4
Activation function	ReLU	ReLU	ReLU	ReLU
Dropout	0	0	0	0
Number of hidden layers	1	1	3	3
Number of hidden neurons	64	64	102	102
Learning rate	0.07	0.07	0.05	0.05
Epoch of RBM	1000	2000	1000	2000