Table 5 Hyper-parameters of ML models adjusted in this study
Algorithm | Hyper-parameters | Explanation | Grid search values |
|---|---|---|---|
Random Forest (RF) | n_estimators | Number of trees in a forest | 100, 150 |
max_depth | Highest depth of the tree | 10, 15 | |
Support Vector Machine (SVM) | C | Penalty parameter | 0.1, 1, 10 |
gamma | Bandwidth parameter | 0.01, 0.1, 1 | |
kernel | Kernel function | RBF | |
LightGBM (LGBM) | n_estimators | Number of trees in a forest | 100, 150 |
learning_rate | Learning rate | 0.01, 0.1 | |
num_leaves | Number of leaves in one tree | 31, 50 | |
max_depth | Highest depth of the tree | −1, 10 | |
XGBoost (XGB) | n_estimators | Number of trees in a forest | 100, 150 |
learning_rate | Learning rate | 0.01, 0.1 | |
max_depth | Highest depth of the tree | 3, 5 | |
CatBoost (CB) | iterations | Number of boosting iterations | 100, 150 |
learning_rate | Learning rate | 0.01, 0.1 | |
depth | Depth of the tree | 4, 6 |
