Optuna Backend

Hyperactive provides wrappers for Optuna’s optimization algorithms, allowing you to use Optuna’s powerful samplers with Hyperactive’s interface.

Note

Optuna must be installed separately:

pip install optuna
# or
pip install hyperactive[all_extras]

Sampler Examples

Sampler	Example
TPE (Tree-Parzen Estimator)	tpe_sampler_example.py
CMA-ES	cmaes_sampler_example.py
Gaussian Process	gp_sampler_example.py
NSGA-II	nsga_ii_sampler_example.py
NSGA-III	nsga_iii_sampler_example.py
QMC (Quasi-Monte Carlo)	qmc_sampler_example.py
Random	random_sampler_example.py
Grid	grid_sampler_example.py

When to Use Optuna Backend

The Optuna backend is useful when you need:

• Multi-objective optimization (NSGA-II, NSGA-III)

• Advanced sampling strategies like CMA-ES or QMC

• Optuna’s pruning capabilities for early stopping

• Compatibility with existing Optuna workflows