Contents

MultiLabelClassificationMetrics.ThresholdSelectionStrategy.recall(_:minimumPrecision:)

A confidence threshold strategy for a recall precision that has at least a minimum precision value.

Declaration

case recall(Float, minimumPrecision: Float)

Discussion

This strategy selects a threshold for each label by searching for the specified recall value on the label’s precision-recall curve. At the recall, the precision must be greater than or equal to the minimum precision value, otherwise a NaN threshold for the corresponding label is returned.

Use this strategy to reduce the rate of false-negative predictions while constraining the false-positive predictions.

See Also

Selection strategies