BOCDConfig

BOCDConfig#

class frouros.detectors.concept_drift.streaming.change_detection.BOCDConfig(model: BaseBOCDModel | None = None, hazard: float = 0.01, min_num_instances: int = 30)#

BOCD (Bayesian Online Change Detection) [adams2007bayesian] configuration.

Parameters:

  • model (Optional[BaseBOCDModel]) – BOCD model, defaults to None. If None, frouros.detectors.concept_drift.streaming.change_detection.bocd.GaussianUnknownMean is used.

  • hazard (float) – hazard value, defaults to 0.01

  • min_num_instances (int) – minimum numbers of instances to start looking for changes, defaults to 30

References:

[adams2007bayesian]

Adams, Ryan Prescott, and David JC MacKay. “Bayesian online changepoint detection.” arXiv preprint arXiv:0710.3742 (2007).

model_type#

alias of GaussianUnknownMean

property model: BaseBOCDModel#

Get model.

Returns:

model

Return type:

BaseBOCDModel

property min_num_instances: int#

Minimum number of instances property.

Returns:

minimum number of instances to start looking for changes

Return type:

int

Contents