smac.multi_objective¶
Interfaces¶
- class smac.multi_objective.AbstractMultiObjectiveAlgorithm[source]¶
Bases:
ABC
A general interface for multi-objective optimizer, depending on different strategies.
- abstract __call__(values)[source]¶
Transform a multi-objective loss to a single loss.
- Parameters:
values (list[float]) – Normalized values in the range [0, 1].
- Returns:
cost – Combined cost.
- Return type:
float
- property meta: dict[str, Any]¶
Returns the meta data of the created object.
- class smac.multi_objective.MeanAggregationStrategy(scenario, objective_weights=None)[source]¶
Bases:
AbstractMultiObjectiveAlgorithm
A class to mean-aggregate multi-objective costs to a single cost.
- Parameters:
scenario (Scenario)
objective_weights (list[float] | None, defaults to None) – Weights for an weighted average. Must be of the same length as the number of objectives.
- __call__(values)[source]¶
Transform a multi-objective loss to a single loss.
- Parameters:
values (list[float]) – Normalized values in the range [0, 1].
- Returns:
cost – Combined cost.
- Return type:
float
- property meta: dict[str, Any]¶
Returns the meta data of the created object.
- class smac.multi_objective.ParEGO(scenario, rho=0.05, seed=None)[source]¶
Bases:
AbstractMultiObjectiveAlgorithm
ParEGO implementation based on https://www.cs.bham.ac.uk/~jdk/UKCI-2015.pdf.
- Parameters:
scenario (Scenario)
rho (float, defaults to 0.05) – A small positive value.
seed (int | None, defaults to None)
- __call__(values)[source]¶
Transform a multi-objective loss to a single loss.
- Parameters:
values (list[float]) – Normalized values in the range [0, 1].
- Returns:
cost – Combined cost.
- Return type:
float
- property meta: dict[str, Any]¶
Returns the meta data of the created object.