Module selection_strategy 

Active learning strategies for intelligent candidate selection

This module provides strategies for selecting which candidates to present to users for evaluation. Instead of random or sequential selection, active learning strategies prioritize candidates that will provide the most useful information for ranking.

Available Strategies

• Sequential: Default behavior - simple sequential/round-robin selection
• UncertaintySampling: Prioritize candidates with highest uncertainty
• ExpectedInformationGain: Select pairs that maximize information gain
• CoverageAware: Balance coverage requirements with exploration

Example

use fugue_evo::interactive::selection_strategy::SelectionStrategy;

// Use uncertainty sampling with coverage bonus
let strategy = SelectionStrategy::UncertaintySampling {
    uncertainty_weight: 1.0,
};

let selected = strategy.select_batch(&candidates, &aggregator, 4);

Enums

SelectionStrategy

Active learning selection strategy