SLEADE#

class capymoa.ssl.SLEADE[source]#

Bases: MOAClassifierSSL

Semi-supervised SLEADE ensemble.

SLEADE method handles partially labelled data and unsupervised drift detection.

>>> from capymoa.ssl import SLEADE
>>> from capymoa.datasets import ElectricityTiny
>>> from capymoa.evaluation import prequential_evaluation
>>>
>>> stream = ElectricityTiny()
>>> clf = SLEADE(stream.get_schema())
>>> results = prequential_evaluation(stream, clf, max_instances=1000)
>>> print(f"{results['cumulative'].accuracy():.1f}")
90.7
__init__(
schema: Schema | None = None,
random_seed: int = 0,
base_ensemble: str = 'StreamingRandomPatches',
confidence_strategy: str = 'ArgMax',
enable_random_threshold: bool = False,
auto_weight_shrinkage: str = 'LabeledNoWarmupDivTotal',
ssl_strategy: str = 'PseudoLabelCheckConfidence',
ssl_min_confidence: float = 0.0,
weight_function: str = 'ConfidenceWeightShrinkage',
pairing_function: str = 'MajorityTrainsMinority',
ssl_weight_shrinkage: float = 100.0,
use_unsupervised_drift_detection: bool = False,
student_learner_for_unsupervised_drift_detection: str = 'trees.HoeffdingTree -g 50 -c 0.01',
drift_detection_method: str = 'ADWINChangeDetector -a 1.0E-5',
unsupervised_detection_weight_window: int = 20,
labeled_window_limit: int = 100,
)[source]#

Construct the SLEADE semi-supervised ensemble.

Parameters:

  • schema – Stream schema.

  • random_seed – Random seed.

  • base_ensemble – Base ensemble learner (e.g., StreamingRandomPatches).

  • confidence_strategy – Confidence strategy (‘Sum’ or ‘ArgMax’).

  • enable_random_threshold – Use random min-confidence threshold.

  • auto_weight_shrinkage – Strategy for automatic weight shrinkage.

  • ssl_strategy – Semi-supervised learning strategy.

  • ssl_min_confidence – Minimum confidence to accept pseudo-label.

  • weight_function – Function for weighting pseudo-labelled instances.

  • pairing_function – Learner pairing function.

  • ssl_weight_shrinkage – Pseudo-label weight shrinkage value.

  • use_unsupervised_drift_detection – Whether to enable unsupervised drift detection.

  • student_learner_for_unsupervised_drift_detection – Student model for drift detection.

  • drift_detection_method – Drift detection algorithm and parameters.

  • unsupervised_detection_weight_window – Window size for unsupervised drift detection weighting.

  • labeled_window_limit – Maximum number of labelled instances in buffer.

cli_help()[source]#
predict(instance: Instance) → int | None[source]#

Predict the label of an instance.

The base implementation calls predict_proba() and returns the label with the highest probability.

Parameters:

instance – The instance to predict the label for.

Returns:

The predicted label or None if the classifier is unable to make a prediction.

predict_proba(
instance,
) → ndarray[tuple[Any, ...], dtype[float64]] | None[source]#

Return probability estimates for each label.

Parameters:

instance – The instance to estimate the probabilities for.

Returns:

An array of probabilities for each label or None if the classifier is unable to make a prediction.

train(instance)[source]#

Train the classifier with a labeled instance.

Parameters:

instance – The labeled instance to train the classifier with.

train_on_unlabeled(instance: Instance)[source]#
random_seed: int#

The random seed for reproducibility.

When implementing a classifier ensure random number generators are seeded.

schema: Schema#

The schema representing the instances.