Enum AcquisitionStrategy 

pub enum AcquisitionStrategy {
    UncertaintySampling,
    ExpectedInformationGain,
    DiversitySampling {
        num_clusters: usize,
    },
    UncertaintyDiversity {
        alpha: f64,
    },
    QueryByCommittee,
    ExpectedModelChange,
}

Active learning acquisition strategies

These strategies determine how samples are selected for labeling based on different criteria such as uncertainty, information gain, or diversity.

Variants

UncertaintySampling

Select samples with highest uncertainty

Selects samples where the model is most uncertain about predictions. This is the most common active learning strategy.

ExpectedInformationGain

Select samples that maximize expected information gain

Chooses samples that are expected to provide the most information about the underlying data distribution.

DiversitySampling

Select diverse samples using clustering

Ensures diversity in selected samples by partitioning into clusters and sampling from each cluster.

Arguments

• num_clusters - Number of clusters to create for diversity

Fields

num_clusters: usize

UncertaintyDiversity

Combine uncertainty and diversity

Balances between uncertain samples and diverse samples using a weighted combination approach.

Arguments

• alpha - Weight for uncertainty vs diversity (0.0-1.0)
  • 1.0 = pure uncertainty sampling
  • 0.0 = pure diversity sampling

Fields

alpha: f64

QueryByCommittee

Query by committee (variance across models)

Selects samples where multiple models disagree the most. Requires ensemble predictions or committee of models.

ExpectedModelChange

Expected model change

Selects samples that are expected to cause the largest change in model parameters when added to training set.

Trait Implementations

impl Clone for AcquisitionStrategy

fn clone(&self) -> AcquisitionStrategy

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for AcquisitionStrategy

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for AcquisitionStrategy

fn default() -> Self

Returns the “default value” for a type.

impl PartialEq for AcquisitionStrategy

fn eq(&self, other: &AcquisitionStrategy) -> bool

Tests for self and other values to be equal, and is used by ==.

fn ne(&self, other: &Rhs) -> bool

Tests for !=. The default implementation is almost always sufficient, and should not be overridden without very good reason.

impl StructuralPartialEq for AcquisitionStrategy