Table 4 The hyperparameters of each model and a comparison of the evaluation metrics
TB | CB | RB | EPB | ||
---|---|---|---|---|---|
Hyperparameter | max_depth | 3 | 3 | 3 | 3 |
learning_rate | 0.02 | 0.01 | 0.02 | 0.02 | |
min_child_weight | 1 | 3 | 2 | 4 | |
colsample_bytree | 0.7 | 0.8 | 0.7 | 0.7 | |
alpha | 0 | 0 | 0 | 0 | |
subsample | 0.8 | 0.8 | 0.8 | 0.8 | |
lambda | 1 | 1 | 1 | 0 | |
random_state | 30 | 30 | 30 | 30 | |
Evaluation metrics | R2 | 0.901 | 0.9137 | 0.941 | 0.913 |
RMSE | 0.145 | 0.125 | 0.120 | 0.120 | |
MAPE | 0.026 | 0.017 | 0.021 | 0.032 |