Research Article
Evolution of Deep Neural Network Architecture Using Particle Swarm Optimization to Improve the Performance in Determining the Friction Angle of Soil
Table 4
Initialization parameters of DNN model.
| Parameters | Value and description |
| Number of hidden layers | 2 | Neurons structure | (25, 25) | Use early stopping | True | Activation function | Tanh | Cost function | MSE | Data used | Training/validation/testing data set | Number of runs (shuffle input data) | 300 |
|
|