sed_eval.metric.equal_error_rate

sed_eval.metric.equal_error_rate(y_true, y_score, eps=2.220446049250313e-16)[source]

Equal error rate (EER)

EER is calculated from the curve of the false negative rate versus the false positive rate. Implementation is based on https://github.com/pafoster/dcase2016_task4/blob/master/evaluation_scripts/eer.py

Parameters:

y_true : numpy.array or list

True binary labels in range {0, 1} or {-1, 1}.

y_score : numpy.array or list

Target scores, can either be probability estimates of the positive class or confidence values.

eps : float

Minimum difference considered equal. Default value numpy.spacing(1)

Returns:

float