Struct ExpectedModelChange 

pub struct ExpectedModelChange { /* private fields */ }

Expected Model Change for Active Learning

Expected Model Change selects samples that are expected to cause the largest change in the model parameters when added to the training set. This strategy looks ahead to predict which samples would be most informative for model updating.

Parameters

• n_samples - Number of samples to select for labeling
• approximation_method - Method for approximating model change (“gradient_norm”, “fisher_information”, “parameter_variance”)
• learning_rate - Learning rate for gradient approximation
• epsilon - Small value for numerical stability
• normalize_scores - Whether to normalize change scores

Examples

use sklears_semi_supervised::ExpectedModelChange;


let X = array![[1.0, 2.0], [2.0, 3.0], [3.0, 4.0]];
let gradients = array![[0.1, 0.2], [0.3, 0.1], [0.05, 0.4]];

let emc = ExpectedModelChange::new()
    .approximation_method("gradient_norm".to_string())
    .n_samples(2);
let selected = emc.select_samples(&X.view(), &gradients.view()).unwrap();

Implementations

impl ExpectedModelChange

pub fn new() -> Self

Create a new ExpectedModelChange instance

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

Set the number of samples to select

pub fn approximation_method(self, method: String) -> Self

Set the approximation method for model change

pub fn learning_rate(self, lr: f64) -> Self

Set the learning rate for gradient approximation

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

Set epsilon for numerical stability

pub fn normalize_scores(self, normalize: bool) -> Self

Set whether to normalize scores

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

Set diversity weight for selection

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

Set batch size for computation

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

Select samples based on expected model change

Trait Implementations

impl Clone for ExpectedModelChange

fn clone(&self) -> ExpectedModelChange

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for ExpectedModelChange

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

Formats the value using the given formatter.

impl Default for ExpectedModelChange

fn default() -> Self

Returns the “default value” for a type.