Table 5 Model parameter settings.
From: Hydropower station scheduling with ship arrival prediction and energy storage
Parameter | Value | Description |
|---|---|---|
n_estimators | 500 | Maximum number of iterations of the weak learner |
learning_rate | 0.05 | Control the iteration rate to prevent over-fitting |
gamma | 0 | The minimum descent value of the loss function required for node splitting, the larger the parameter value, the more conservative the algorithm |
\(max\_depth\) | 9 | The maximum depth of the tree, the larger the value the more complex the model. Overfitting can be controlled by this value |
alpha | 1 | L1 regularization term for the weights |
lambda | 0.5 | L2 regularization term for the weights |
subsample | 1 | Control the proportion of random sampling for each tree |
booster | gbtree | Select base classifier, specify ascent model, commonly tree or linear model |
