Skip to main content

Table 1 The tuning parameter values of SMOTE-based machine learning methods

From: Hospital mortality prediction in traumatic injuries patients: comparing different SMOTE-based machine learning algorithms

Methods

Hyperparameters

Definition

Value

ANN

Size

The number of nodes in the hidden layer

5

Weight decay

The regularization parameter to avoid overfitting

0.1

SVM

Gamma

The width of the radial basis function kernel

0.12

Cost

The parameter that controls the complexity of the model

1

RF

mtry

Number of variables randomly selected as candidates for each tree

2

ntree

The number of trees

500

DT

minsplit

The minimum number of observations in a node

10

minbucket

The minimum number of observations in any terminal node

3

XGBoost

nrounds

The maximum number of iterations

100

eta

Learning rate

0.3

gamma

Regularization parameter to prevent overfitting

5

max depth

The depth of the tree

3