LeastConfidentSampling

class probly.evaluation.active_learning.strategies.LeastConfidentSampling[source]

Bases: object

Selects unlabeled samples where the model is least confident in its top prediction.

A low maximum probability indicates the model is uncertain overall, making these samples informative to label.

select(estimator: Estimator, pool: ActiveLearningPool, n: int) → ArrayLike[source]

Return n indices with the lowest maximum class probability.