Active Learning

Contains the implementations for active learning methods

Implements the supported active learning methods.

class calisim.active_learning.implementation.ActiveLearningMethod(calibration_func: Callable, specification: ActiveLearningMethodModel, engine: str = 'skactiveml', implementation: type[CalibrationWorkflowBase] | CalibrationWorkflowBase | None = None)

The active learning method class.

pydantic model calisim.active_learning.implementation.ActiveLearningMethodModel

The active learning method data model.

Parameters:

BaseModel (CalibrationModel) – The calibration base model class.

Config:

• arbitrary_types_allowed: bool = True

• protected_namespaces: tuple = ()

Fields:

• query_strategy (str)

field query_strategy: str = 'greedy_sampling_target'

The active learning query strategy

calisim.active_learning.implementation.get_implementations() → dict[str, str]

Get the calibration implementations for active learning.

Returns:

The dictionary

of calibration implementations for active learning.

Return type:

Dict[str, str]

scikit-activeml

Contains the implementations for active learning methods using scikit-activeml

Implements the supported active learning methods using the scikit-activeml library.

class calisim.active_learning.skactiveml_wrapper.SkActiveMLActiveLearning(calibration_func: Callable, specification: CalibrationModel, task: str)

The scikit-activeml active learning method class.

analyze() → None

Analyze the results of the simulation calibration procedure.

execute() → None

Execute the simulation calibration procedure.