saealib.SurrogateAccuracyMetric#

class saealib.SurrogateAccuracyMetric[source]#

Bases: ABC

Abstract base for a scalar accuracy metric.

Methods

__init__

compute

Compute the metric.

Method Details

SurrogateAccuracyMetric.__init__()#
abstractmethod SurrogateAccuracyMetric.compute(y_true, y_pred)[source]#

Compute the metric.

Parameters:

  • y_true (np.ndarray) – True objective values. shape: (n_samples, n_obj)

  • y_pred (np.ndarray) – Predicted objective values. Same shape as y_true.

Returns:

Scalar metric value. Interpretation (higher/lower = better) depends on the concrete metric.

Return type:

float