optuna.samplers.nsgaii.PolynomialMutation

class optuna.samplers.nsgaii.PolynomialMutation(eta=20.0)[source]

Polynomial mutation operation used by NSGAIISampler.

This operator mutates a real-valued parameter according to the polynomial probability distribution.

This implementation follows the polynomial mutation procedure used in the revision 1.1.6 of the original NSGA-II C implementation released as Multi-objective NSGA-II code in C.

Deb, K., Pratap, A., Agarwal, S. and Meyarivan, T. A fast and elitist multiobjective genetic algorithm: NSGA-II. IEEE Transactions on Evolutionary Computation, 6(2), 182-197 (2002).
Multi-objective NSGA-II code in C, Revision 1.1.6

Parameters:: eta (float) – Distribution index. Larger values make mutated parameter values closer to the original value.

Note

Added in v5.0.0 as an experimental feature. The interface may change in newer versions without prior notice. See https://github.com/optuna/optuna/releases/tag/v5.0.0.

Methods

mutation(param, rng, study, search_space_bounds)

Mutate the given parameter.

mutation(param, rng, study, search_space_bounds)[source]

Mutate the given parameter.

Parameters:

param (float) – A parameter value in the transformed numerical search space.
rng (np.random.RandomState) – An instance of numpy.random.RandomState.
study (Study) – Target study object.
search_space_bounds (np.ndarray) – A numpy.ndarray with shape (2,) representing the numerical distribution bounds constructed from transformed search space.

Returns:

A mutated parameter value in the transformed numerical search space.

Return type:

float