Scikit-Learn Ensemble Methods
The following Scikit-Learn module types are ensemble methods.
Modules
- AdaBoostClassifier
- AdaBoostRegressor
- BaggingClassifier
- BaggingRegressor
- ExtraTreesClassifier
- ExtraTreesRegressor
- GradientBoostingClassifier
- GradientBoostingRegressor
- HistGradientBoostingClassifier
- HistGradientBoostingRegressor
- IsolationForest
- RandomForestClassifier
- RandomForestRegressor
- RandomTreesEmbedding
AdaBoostClassifier
Scikit-Learn module documentation: AdaBoostClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of estimators to train | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Learning rate for boosting | --skl_shrinkage | Float > 0 | None | learning_rate |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
AdaBoostRegressor
Scikit-Learn module documentation: AdaBoostRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of estimators to train | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Learning rate for boosting | --skl_shrinkage | Float > 0 | None | learning_rate |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
BaggingClassifier
Scikit-Learn module documentation: BaggingClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of estimators to train | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Fraction of samples to draw per estimator | --skl_max_samples | Float in (0, 1] | None | max_samples |
| Number of features to consider for each estimator | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
BaggingRegressor
Scikit-Learn module documentation: BaggingRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of estimators to train | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Fraction of samples to draw per estimator | --skl_max_samples | Float in (0, 1] | None | max_samples |
| Number of features to consider for each estimator | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
ExtraTreesClassifier
Scikit-Learn module documentation: ExtraTreesClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the forest | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Quality measure for splits | --skl_criterion | gini | entropy | log_loss | None | criterion |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Minimum samples required to split a node | --skl_min_samples_split | Integer ≥ 2 | None | min_samples_split |
| Minimum samples required at a leaf node | --skl_min_samples_leaf | Integer ≥ 1 | None | min_samples_leaf |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
ExtraTreesRegressor
Scikit-Learn module documentation: ExtraTreesRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the forest | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Quality measure for splits | --skl_criterion | squared_error | friedman_mse | None | criterion |
absolute_error | poisson | ||||
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Minimum samples required to split a node | --skl_min_samples_split | Integer ≥ 2 | None | min_samples_split |
| Minimum samples required at a leaf node | --skl_min_samples_leaf | Integer ≥ 1 | None | min_samples_leaf |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
GradientBoostingClassifier
Scikit-Learn module documentation: GradientBoostingClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of boosting stages | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Learning rate (shrinks each tree’s contribution) | --skl_shrinkage | Float > 0 | None | learning_rate |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Fraction of samples used per stage | --skl_subsample | Float in (0, 1] | None | subsample |
| Quality measure for splits | --skl_criterion | friedman_mse | squared_error | None | criterion |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Convergence tolerance | --skl_tol | Float ≥ 0 | None | tol |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
GradientBoostingRegressor
Scikit-Learn module documentation: GradientBoostingRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of boosting stages | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Learning rate (shrinks each tree’s contribution) | --skl_shrinkage | Float > 0 | None | learning_rate |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Fraction of samples used per stage | --skl_subsample | Float in (0, 1] | None | subsample |
| Quality measure for splits | --skl_criterion | friedman_mse | squared_error | None | criterion |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Convergence tolerance | --skl_tol | Float ≥ 0 | None | tol |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
HistGradientBoostingClassifier
Scikit-Learn module documentation: HistGradientBoostingClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Learning rate (shrinks each tree’s contribution) | --skl_shrinkage | Float > 0 | None | learning_rate |
| Maximum number of iterations | --skl_max_iter | Integer ≥ 1 | None | max_iter |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Convergence tolerance | --skl_tol | Float ≥ 0 | None | tol |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
HistGradientBoostingRegressor
Scikit-Learn module documentation: HistGradientBoostingRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Learning rate (shrinks each tree’s contribution) | --skl_shrinkage | Float > 0 | None | learning_rate |
| Maximum number of iterations | --skl_max_iter | Integer ≥ 1 | None | max_iter |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Convergence tolerance | --skl_tol | Float ≥ 0 | None | tol |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
IsolationForest
Scikit-Learn module documentation: IsolationForest
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the ensemble | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Fraction of samples to draw per tree | --skl_max_samples | Float in (0, 1] | None | max_samples |
| Proportion of outliers in the dataset | --skl_contamination | Float in (0, 0.5] | None | contamination |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
RandomForestClassifier
Scikit-Learn module documentation: RandomForestClassifier
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the forest | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Quality measure for splits | --skl_criterion | gini | entropy | log_loss | None | criterion |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Minimum samples required to split a node | --skl_min_samples_split | Integer ≥ 2 | None | min_samples_split |
| Minimum samples required at a leaf node | --skl_min_samples_leaf | Integer ≥ 1 | None | min_samples_leaf |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
RandomForestRegressor
Scikit-Learn module documentation: RandomForestRegressor
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the forest | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Quality measure for splits | --skl_criterion | squared_error | friedman_mse | None | criterion |
absolute_error | poisson | ||||
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Minimum samples required to split a node | --skl_min_samples_split | Integer ≥ 2 | None | min_samples_split |
| Minimum samples required at a leaf node | --skl_min_samples_leaf | Integer ≥ 1 | None | min_samples_leaf |
| Number of features to consider for best split | --skl_max_features | sqrt | log2 | None | max_features |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |
RandomTreesEmbedding
Scikit-Learn module documentation: RandomTreesEmbedding
Arguments
| Description | Zero2Neuro Argument | Possible Values | Default Value | Scikit-Learn Parameter |
|---|---|---|---|---|
| Number of CPU cores to use | --cpus_per_task | Integer ≥ 1 | None | n_jobs |
| Number of trees in the forest | --skl_n_estimators | Integer ≥ 1 | None | n_estimators |
| Maximum depth of each tree | --skl_max_depth | Integer ≥ 1 | None | max_depth |
| Random seed for reproducibility | --skl_random_state | Integer | None | random_state |