Table 1 The hyperparameters for proposed six machine learning models
Model | Hyperparameters |
|---|---|
AdaBoost | n_estimators:10 |
| Â | learning_rate:1.0 |
| Â | algorithm: SAMME.R |
| Â | base_estimator: deprecated |
LR | penalty:none |
| Â | dual:False |
| Â | tol:1e-4 |
| Â | c:1.0 |
| Â | fit_intercept: True |
| Â | intercept_scaling:1.0 |
| Â | class_weight: None |
| Â | solver:lbfgs |
| Â | max_iter:100 |
| Â | verbose:0 |
| Â | warm_start: False |
| Â | n_jobs: None |
MLP | hidden_layer_sizes:100 |
| Â | activation: relu |
| Â | solver: lbfgs |
| Â | alpha:0.0001 |
| Â | learning_rate: constant |
| Â | learning_rate_init:0.01 |
| Â | power_t:0.5 |
| Â | max_iter:200 |
| Â | shuffle: True |
Bagging | n_estimators:10 |
| Â | bootstrap: True |
| Â | bootstrap_features: False |
| Â | oob_score: False |
| Â | warm_start: False |
| Â | n_jobs: None |
| Â | verbose:0 |
| Â | base_estimator: deprecated |
| Â | max_samples:0.5 |
| Â | max_features:0.5 |
GBM | n_estimators:100 |
| Â | learning_rate:1.0 |
| Â | max_depth:1.0 |
| Â | subsample:1.0 |
| Â | criterion: friedman_mse |
| Â | min_samples_split:2 |
| Â | min_samples_leaf:1 |
| Â | min_weight_fraction_leaf:0.0 |
| Â | min_impurity_decrease:0.0 |
| Â | init: None |
| Â | max_features: None |
| Â | verbose:0 |
| Â | max_leaf_nodes: None |
| Â | warm_start: False |
| Â | validation_fraction:0.1 |
| Â | n_iter_no_change: None |
| Â | tol:1e-4 |
| Â | ccp_alpha:0.0 |
XGBoost | n_estimators:360 |
| Â | max_depth:1 |
| Â | learning_rate:1.6 |