|
| Parameters | Descriptions and settings |
|
| n_estimators | The number of classifiers, also called the number of iterations, is the number of decision trees in the forest |
| Criterion | The standard used for splitting nodes, the default is gini |
| Max_depth | Maximum depth of trees |
| Min_samples_split | Minimum number of samples required to split a node inside the tree, defaults to 2 |
| Min_samples_leaf | Minimum number of samples required at leaf nodes, defaults to 1 |
| Min_weight_fraction_leaf | Minimum weighted score in the sum of weights at all leaf nodes, defaults to 0 |
| Max_features | The number of features to consider when finding the best segmentation, the default is none; that is, all features are considered |
| Max_leaf_nodes | The maximum number of leaf nodes, which must be an integer, the default is none |
| Min_impurity_decrease | If the decrement of the split index is greater than this value, then split, default is 0 |
| Bootstrap | Whether there is a randomly selected sample to put back, the default is true |
| Oob_score | Whether to use out-of-bag samples to evaluate the quality of the model, set to true |
| N_jobs | The number of parallel calculations, the default is none |
| Random_state | Controls the randomness of bootstrap and randomness of the selected samples. In order to facilitate the adjustment of other parameters, this parameter adopts the default value |
| Verbose | Controls verbosity when fitting and predicting, default is 0 |
| Warm_start | Whether to use the trained model and add more base learners to it, set to false |
|
