Table A1. Prediction algorithms with... | Emerald Publishing

Table A1.

Prediction algorithms with evaluated hyperparameter range

Model	Evaluated hyperparameter range (same for D1 and D2 datasets)
AdaBoost	“n_estimators”: [50, 500] “learning_rate”: [0.01, 1.0] “max_depth”: [2, 10] “Loss”: {“linear”, “square”, “exponential”}
CatBoost	“iteration”: [50, 1000] “learning_rate”: [0.01, 0.3] “depth”: [4, 10] “l2_leaf_reg”: [1, 10] “subsample”: [0.5, 0.95] “random_strength”: [0.0, 1.0]
Extra tree	“n_estimators”: [50, 500] “max_depth”: [2, 15] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “max_features”: {“sqrt”, “log2”, 1.0, 0.2, 0.5, 0.8} “Criterion”: {“squared_error”, “absolute_error”}
Gradient boosting	“n_estimators”: [50, 500] “learning_rate”: [0.01, 0.3] “max_depth”: [2, 8] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “subsample”: [0.5, 1.0] “Loss”: {“squared_error”, “absolute_error”, “huber”}
Random forest	“n_estimators”: [50, 500] “max_depth”: [2, 15] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “max_features”: {“sqrt”, “log2”, 1.0} “Criterion”: {“squared_error”, “absolute_error”} “Bootstrap”: {true, false}
XGBoost	“learning_rate”: [0.01, 1.0] “max_depth”: [2, 10] “n_estimators”: [50, 500] “gamma”: [50, 500] “subsample”: [0.5, 1.0] “colsample_bytree”: [0.5, 1.0]
Slime mould algorithm	“epoch”: 500 “pop_size”: 30 “probability_threshold”: 0.03

Model	Evaluated hyperparameter range (same for D1 and D2 datasets)
AdaBoost	“n_estimators”: [50, 500] “learning_rate”: [0.01, 1.0] “max_depth”: [2, 10] “Loss”: {“linear”, “square”, “exponential”}
CatBoost	“iteration”: [50, 1000] “learning_rate”: [0.01, 0.3] “depth”: [4, 10] “l2_leaf_reg”: [1, 10] “subsample”: [0.5, 0.95] “random_strength”: [0.0, 1.0]
Extra tree	“n_estimators”: [50, 500] “max_depth”: [2, 15] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “max_features”: {“sqrt”, “log2”, 1.0, 0.2, 0.5, 0.8} “Criterion”: {“squared_error”, “absolute_error”}
Gradient boosting	“n_estimators”: [50, 500] “learning_rate”: [0.01, 0.3] “max_depth”: [2, 8] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “subsample”: [0.5, 1.0] “Loss”: {“squared_error”, “absolute_error”, “huber”}
Random forest	“n_estimators”: [50, 500] “max_depth”: [2, 15] “min_samples_split”: [2, 20] “min_samples_leaf”: [1, 10] “max_features”: {“sqrt”, “log2”, 1.0} “Criterion”: {“squared_error”, “absolute_error”} “Bootstrap”: {true, false}
XGBoost	“learning_rate”: [0.01, 1.0] “max_depth”: [2, 10] “n_estimators”: [50, 500] “gamma”: [50, 500] “subsample”: [0.5, 1.0] “colsample_bytree”: [0.5, 1.0]
Slime mould algorithm	“epoch”: 500 “pop_size”: 30 “probability_threshold”: 0.03