Table 4 ANN hyperparameter ranges employed for grid optimization and final ANNs.

From: Smart extrusion via data-driven prediction of grain size and peripheral coarse grain defect formation

 

Grid optimization

Final parameters,

PCG formation

Final parameters,

grain size

Activation function for fully connected layers

Rectified linear unit (ReLU); Hyperbolic tangent (tanh);

Sigmoid

ReLU

Sigmoid

Number of fully connected layers

1–2

1

2

Number of neurons per layer

1–50

Layer 1: 22

Layer 1: 20

Layer 2: 20

Back to article page