Table 2 The parameters of the eight candidate models

DT

max_depth = 10, min_samples_leaf = 20, min_samples_split = 300, random_state = 1

SVM

kernel = linear, C = 0.001, tol = 0.0001

RF

bootstrap = True, max_depth = 5, n_estimators = 50, random_state = 1

ET

bootstrap = False, max_depth = 3, n_estimators = 50, random_state = 1

GBDT

learning_rate = 0.05, max_depth = 3, n_estimators = 50, subsample = 0.6, random_state = 1

ADB

base_estimator = DecisionTreeClassifier (max_depth = 3), learning_rate = 0.01, n_estimators = 100, random_state = 1

Bagging

base_estimator = DecisionTreeClassifier (max_depth = 5), n_estimators = 300, bootstrap = True, max_features = 0.6, max_samples = 0.6, random_state = 1

XGB

learning_rate = 0.01, max_depth = 5, n_estimators = 200, subsample = 0.6, colsample_bytree = 0.8, random_state = 1