Skip to content

[FEATURE] Add support for outlier detectors #162

@asdf32768

Description

@asdf32768

Hi,
I'm trying to finetune Isolation Forest hyperparameters (
param_grid = {'contamination': Continuous(0.001, 0.5, distribution='log-uniform'), 'n_estimators': Integer(100, X_train.shape[0]), 'max_samples': Integer(1, X_train.shape[0]), 'max_features': Integer(1, X_train.shape[1]), 'boostrap': Categorical([True, False])})

using GASearchCV and I then get this error:

272 # Check that the estimator is compatible with scikit-learn
273 if not is_classifier(self.estimator) and not is_regressor(self.estimator):
--> 274 raise ValueError(f"{self.estimator} is not a valid Sklearn classifier or regressor")
275
276 if criteria not in Criteria.list():
ValueError: IsolationForest() is not a valid Sklearn classifier or regressor

Could you please add support for anomaly detection algorithms, such as IsolationForest.

Metadata

Metadata

Assignees

No one assigned

    Labels

    new featureDescribe the request of new features

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions