KSWINConfig

KSWINConfig#

class frouros.detectors.concept_drift.streaming.window_based.KSWINConfig(alpha: float = 0.0001, seed: int | None = None, min_num_instances: int = 100, num_test_instances: int = 30)#

KSWIN (Kolmogorov-Smirnov Windowing) [raab2020reactive] configuration.

Parameters:

  • alpha (float) – significance value, defaults to 0.0001

  • seed (Optional[int]) – seed value, defaults to None

  • min_num_instances (int) – minimum numbers of instances to start looking for changes, defaults to 100

  • num_test_instances (int) – numbers of instances to be used by the statistical test, defaults to 30

Raises:

ValueError – Value error exception if seed is not valid

References:

[raab2020reactive]

Raab, Christoph, Moritz Heusinger, and Frank-Michael Schleif. “Reactive soft prototype computing for concept drift streams.” Neurocomputing 416 (2020): 340-351.

property alpha: float#

Significance value property.

Returns:

significance value

Return type:

float

property num_test_instances: int#

Number of tests instances property.

Returns:

number of tests instances to be used by the statistical test

Return type:

int

property min_num_instances: int#

Minimum number of instances property.

Returns:

minimum number of instances to start looking for changes

Return type:

int

Contents