Module bayesian_opt 

Bayesian optimization with Gaussian process surrogates.

Provides a full Bayesian optimization loop: fit a GP surrogate, evaluate an acquisition function to pick the next candidate, observe the objective, and iterate. Supports RBF, Matérn-5/2, and Periodic kernels.

Structs

BayesOpts

Configuration for BayesianOptimizer.

BayesianOptimizer

Bayesian optimization over a bounded box using a GP surrogate.

GaussianProcess

A Gaussian process regressor using a stationary kernel.

KernelParams

Hyper-parameters for the GP kernel and likelihood noise.

Enums

AcquisitionFn

Acquisition function used to select the next candidate point.

KernelType

The covariance kernel used by the Gaussian process.

Functions

acquisition_value

Evaluate the acquisition function at a single point.

cholesky

Compute the lower-triangular Cholesky factor L such that A = L Lᵀ.

kernel_eval

Evaluate the kernel between two input vectors a and b.

latin_hypercube_sample

Generate a Latin hypercube sample of n points in a dim-dimensional box.