FPR95#

class torch_uncertainty.metrics.classification.FPR95(pos_label, **kwargs)[source]#

Compute the False Positive Rate at 95% Recall.

This is a specific case of the more general FPRx metric, where the recall level is fixed at 95%.

Parameters:
  • pos_label (int) – The positive label (e.g., 1 for OOD samples).

  • kwargs – Additional arguments to pass to the FPRx metric class.

See also

  • FPRx - The base metric that allows customization of the recall level.