incerto.active.BadgeSampling

incerto.active.BadgeSampling#

class incerto.active.BadgeSampling(batch_size=100)[source]#

Bases: object

BADGE (Batch Active learning by Diverse Gradient Embeddings).

Combines gradient-based embeddings with k-MEANS++ for diverse batch selection.

Reference:

Ash et al., “Deep Batch Active Learning by Diverse, Uncertain Gradient Lower Bounds” (ICLR 2020)

Parameters:

batch_size (int) – Number of samples to select

__init__(batch_size=100)[source]#
Parameters:

batch_size (int)

Methods

__init__([batch_size])

query(model, x_unlabeled)

Select batch using BADGE.
__init__(batch_size=100)[source]#
Parameters:

batch_size (int)

query(model, x_unlabeled)[source]#

Select batch using BADGE.

Parameters:

  • model (Module) – Model

  • x_unlabeled (Tensor) – Unlabeled data

Return type:

Tensor

Returns:

Indices of selected samples

Contents