Struct UncertaintySampling 

pub struct UncertaintySampling { /* private fields */ }

Uncertainty Sampling for Active Learning

Uncertainty sampling selects samples for labeling based on the model’s uncertainty about their predictions. It supports multiple uncertainty measures including entropy, margin, and least confident sampling.

Parameters

• strategy - Uncertainty sampling strategy (“entropy”, “margin”, “least_confident”)
• n_samples - Number of samples to select for labeling
• temperature - Temperature scaling for probability calibration
• diversity_weight - Weight for diversity-based selection
• batch_size - Size of batches for efficient computation

Examples

use sklears_semi_supervised::UncertaintySampling;


let probas = array![
    [0.9, 0.1],
    [0.6, 0.4],
    [0.5, 0.5],
    [0.8, 0.2]
];

let us = UncertaintySampling::new()
    .strategy("entropy".to_string())
    .n_samples(2);
let selected_indices = us.select_samples(&probas.view()).unwrap();

Implementations

impl UncertaintySampling

pub fn new() -> Self

Create a new UncertaintySampling instance

pub fn strategy(self, strategy: String) -> Self

Set the uncertainty sampling strategy

pub fn n_samples(self, n_samples: usize) -> Self

Set the number of samples to select

pub fn temperature(self, temperature: f64) -> Self

Set the temperature for probability calibration

pub fn diversity_weight(self, weight: f64) -> Self

Set the diversity weight

pub fn batch_size(self, batch_size: usize) -> Self

Set the batch size for computation

pub fn random_tie_breaking(self, random: bool) -> Self

Enable/disable random tie breaking

pub fn select_samples( &self, probas: &ArrayView2<'_, f64>, ) -> SklResult<Vec<usize>>

Select samples based on uncertainty

Trait Implementations

impl Clone for UncertaintySampling

fn clone(&self) -> UncertaintySampling

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for UncertaintySampling

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

Formats the value using the given formatter.

impl Default for UncertaintySampling

fn default() -> Self

Returns the “default value” for a type.