SMOTEENNConverter

Converter

DashAI.back.converters.imbalanced_learn.SMOTEENNConverter

---

Hybrid sampler combining SMOTE oversampling with Edited Nearest Neighbours cleaning.

SMOTE-ENN is a two-stage resampling strategy for imbalanced classification:

1. Over-sampling — SMOTE generates synthetic minority-class examples by interpolating between each minority sample and its k-nearest minority neighbours, increasing the minority class size.
2. Cleaning — Edited Nearest Neighbours (ENN) removes any sample (from either class) whose class label disagrees with the majority vote of its nearest neighbours, reducing class overlap and borderline noise.

The combined effect is a more balanced and cleaner decision boundary than either technique alone. All schema parameters are forwarded to imbalanced-learn's SMOTEENN and its internal SMOTE sub-estimator.

References

• [1] Chawla, N.V. et al. (2002). "SMOTE: Synthetic Minority Over-sampling Technique." JAIR, 16, 321-357. https://arxiv.org/abs/1106.1813
• [2] Batista, G.E.A.P.A. et al. (2004). "A study of the behaviour of several methods for balancing machine learning training data." ACM SIGKDD Explorations, 6(1), 20-29.
• [3] https://imbalanced-learn.org/stable/references/generated/imblearn.combine.SMOTEENN.html

Parameters

sampling_strategy, default=auto

Sampling strategy to apply SMOTE and clean the dataset.

random_state, default=None

Seed used for reproducibility.

k_neighbors : integer, default=5

Number of neighbors used by SMOTE.

Methods

get_output_type(self, column_name: str = None) -> DashAI.back.types.dashai_data_type.DashAIDataType

Defined on SMOTEENNConverter

Not implemented; type preservation is handled in transform.

Parameters

column_name : str or None, optional

Name of the column whose output type is queried. Ignored because this method always raises. Default None.

changes_row_count(self) -> bool

Defined on ImbalancedLearnWrapper

Return True because all samplers add or remove rows.

Returns

bool

Always True.

fit(self, x: 'DashAIDataset', y: 'DashAIDataset') -> Type[DashAI.back.converters.base_converter.BaseConverter]

Defined on ImbalancedLearnWrapper

Resample the dataset by calling fit_resample and store the result.

Parameters

x : DashAIDataset

The input feature dataset.

y : DashAIDataset

The target label dataset (required; must be non-empty).

Returns

Type[BaseConverter]

The fitted sampler instance (self).

get_metadata(cls) -> 'Dict[str, Any]'

Defined on BaseConverter

Get metadata for the converter, used by the DashAI frontend.

Parameters

cls : type

The converter class (injected automatically by Python for classmethods).

Returns

Dict[str, Any]

Dictionary containing display name, short description, image preview path, category, icon, color, and whether the converter is supervised.

get_schema(cls) -> dict

Defined on ConfigObject

Generates the component related Json Schema.

Returns

dict

Dictionary representing the Json Schema of the component.

transform(self, x: 'DashAIDataset', y: Optional[ForwardRef('DashAIDataset')] = None) -> 'DashAIDataset'

Defined on ImbalancedLearnWrapper

Return the resampled dataset stored during fit.

Parameters

x : DashAIDataset

The original feature dataset (used only for type information).

y : DashAIDataset, optional

The original target dataset (used only for type information). Defaults to None.

Returns

DashAIDataset

The combined resampled dataset (features + target) produced by fit.

validate_and_transform(self, raw_data: dict) -> dict

Defined on ConfigObject

It takes the data given by the user to initialize the model and returns it with all the objects that the model needs to work.

Parameters

raw_data : dict

A dictionary with the data provided by the user to initialize the model.

Returns

dict

A validated dictionary with the necessary objects.