Table 3 Machine learning algorithms and parameters.

FW sub-model w₁: Random Forest

Number of trees

21

Criterion

MSE

Maximal depth

70

FW sub-model w₂: Random Forest

Number of trees

11

Criterion

MSE

Maximal depth

90

FW sub-model w₃: Random Forest

Number of trees

41

Criterion

MSE

Maximal depth

29

T-LA sub-model w₁: Random Forest

number of trees

90

Criterion

MSE

Maximal depth

60

T-LA sub-model w₂: Random Forest

number of trees

60

Criterion

MSE

Maximal depth

21

T-LA sub-model w₃: Random Forest

number of trees

70

Criterion

MSE

Maximal depth

70

N sub-model w₁: Deep Neural Network

Activation function

ReLU

Number of layers

3

Number of neurons

50

Learning rate

0.4

Loss function

MSE

Optimizer

SGD

Epochs

100

N sub-model w₂: Random Forest

Number of trees

21

Criterion

MSE

Maximal depth

29

N sub-model w₃: SVM

Kernel function

ANOVA

C

-1

Gamma

700