PrequentialError

class frouros.metrics.prequential_error.PrequentialError(alpha: int | float = 1.0, name: str | None = None)

Prequential error [dawid1984present] using fading factor [gama2009issues] metric.

Parameters:

• alpha (Union[int, float]) – fading factor value, defaults to 1.0

• name (Optional[str]) – name value, defaults to None. If None, the name will be set to PrequentialError.

References:

[dawid1984present]

Dawid, A. Philip. “Present position and potential developments: Some personal views statistical theory the prequential approach.” Journal of the Royal Statistical Society: Series A (General) 147.2 (1984): 278-290.

[gama2009issues]

Gama, Joao, Raquel Sebastiao, and Pedro Pereira Rodrigues. “Issues in evaluation of stream learning algorithms.” Proceedings of the 15th ACM SIGKDD international conference on Knowledge discovery and data mining. 2009.

Example:

>>> from frouros.metrics import PrequentialError
>>> metric = PrequentialError(alpha=0.9)
>>> X = [1, 1, 0, 1, 0, 0]
>>> Y = [1, 0, 0, 0, 1, 1]
>>> for i, (X_sample, Y_sample) in enumerate(zip(X, Y)):
...     error_value = 1 - (X_sample == Y_sample)
...     prequential_error = metric(error_value=error_value)
...     print(f"Metric={prequential_error:.5f} at step {i}")
Metric=0.00000 at step 0
Metric=0.52632 at step 1
Metric=0.33210 at step 2
Metric=0.52632 at step 3
Metric=0.64199 at step 4
Metric=0.71839 at step 5

property alpha: int | float

Fading factor property.

Returns:

fading factor value

Return type:

Union[int, float]

property cumulative_instances: int | float

Cumulative instances’ property.

Returns:

fading factor value

Return type:

Union[int, float]

property cumulative_fading_error: int | float

Cumulative fading error property.

Returns:

cumulative facing error value

Return type:

Union[int, float]

property name: str

Metrics’ name property.

Returns:

metrics’ name

Return type:

str

reset() → None

Reset method.