Table 2 Set of hyperparameters and testing values chosen for the cross-validation optimization process on each ML model.

DT

Criterion of splitting

{Gini, Entropy, Log-Loss}

Splitter in the nodes

{Best, Random}

Maximum depth

From 1 to 30 with steps of 5

Minimum size of samples per leaf

From 2 to 30 with steps of 5

Minimum size of samples to split

From 1 to 30 with steps of 5

Maximum number of features

{Squared root, logarithmic}

Class weights

{Standard, Balanced}

LR

Penalty

{L-1 norm, L-2 norm}

Size of penalization

From 0.01 to 100 with steps of 0.05

Class weights

{Standard, Balanced}

RF

Number of trees

From 10 to 200 with steps of 50

Criterion of splitting

{Gini, Entropy, Log-Loss}

Maximum depth

From 1 to 20 with steps of 5

Minimum size of samples per leaf

From 2 to 20 with steps of 5

Minimum size of samples to split

From 1 to 20 with steps of 5

Maximum number of features

{Squared root, logarithmic}

Class weights

{Standard, balanced}

XGB

Number of trees

From 10 to 200 with steps of 50

Maximum depth

From 1 to 20 with steps of 5

Subsample ratio

{0.2, 0.5, 0.8, 1.0}

Learning rate

From 0.01 to 1 with steps of 0.2

Subsample ratio of columns

{0.1, 0.2, 0.5, 1.0}

Class weights

{Standard, balanced}

SVM

Kernel

{Linear, RBF}

Control parameter

From 0.01 to 100 with steps of 0.5

Class weights

{Standard, Balanced}