module change_detection.evaluation.measures.missed_detection_rate

Missed Detection Rate Measure.

This function returns the rate of known concept drifts that the model has failed to detect. The measure was introduced by Bifet, Albert, et al. "CD-MOA: Change detection framework for massive online analysis." International Symposium on Intelligent Data Analysis. Springer, Berlin, Heidelberg, 2013.

The missed detection rate is equal to 1 - detected_change_rate.

Copyright (C) 2022 Johannes Haug.

function missed_detection_rate

missed_detection_rate(
    evaluator: float.change_detection.evaluation.change_detection_evaluator.ChangeDetectionEvaluator,
    drifts: list,
    n_delay: int
) → float

Calculates the rate of missed known concept drifts.

The missed detection rate is equal to 1 - the detected change rate.

Args:

  • evaluator: The ChangeDetectionEvaluator object.
  • drifts: List of time steps corresponding to detected concept drifts.
  • n_delay: The number of observations after a known concept drift, during which we count the detections made by the model as true positives.

Returns:

  • float: The fraction of missed known concept drifts.

This file was automatically generated via lazydocs.