Module optimization 

Optimization engine with Bayesian optimization.

Implements Bayesian optimization using Gaussian Process surrogates for sample-efficient black-box optimization (Kaizen: continuous improvement).

Acquisition Functions

• Expected Improvement (EI): Balances exploration and exploitation
• Upper Confidence Bound (UCB): Tunable exploration via kappa
• Probability of Improvement (PI): Conservative improvement strategy

Example

use simular::domains::optimization::{BayesianOptimizer, OptimizerConfig, AcquisitionFunction};

let config = OptimizerConfig {
    bounds: vec![(-5.0, 5.0), (-5.0, 5.0)],
    acquisition: AcquisitionFunction::ExpectedImprovement,
    ..Default::default()
};

let mut optimizer = BayesianOptimizer::new(config);

// Add initial observations
optimizer.observe(vec![0.0, 0.0], 1.5);
optimizer.observe(vec![1.0, 1.0], 0.8);

// Get next suggested point
let next_point = optimizer.suggest();

Structs

BayesianOptimizer

Bayesian optimizer using Gaussian Process surrogate.

GaussianProcess

Gaussian Process surrogate model.

OptimizationResult

Result of optimization run.

OptimizerConfig

Configuration for Bayesian optimizer.

Enums

AcquisitionFunction

Acquisition functions for Bayesian optimization.