SCODCovAt5Risk#

class torch_uncertainty.metrics.classification.SCODCovAt5Risk(ood_cost=0.5, **kwargs)[source]#

Calculate the maximum coverage at 5% SCOD risk.

This is the fixed-threshold variant of SCODCovAtxRisk with \(\tau=0.05\):

\[\operatorname{SCOD\text{-}Cov@5Risk} = \max\left\{ \kappa_k : r(\kappa_k)\leq 0.05,\quad k\in\{1,\ldots,N\} \right\}.\]

If no positive coverage has a SCOD risk at most 5%, the metric returns nan.

Parameters:

  • ood_cost (float) – Relative cost \(c_{\mathrm{OOD}}\) of accepting an OOD sample. The cost of accepting a misclassified ID sample is 1 - ood_cost. Defaults to 0.5.

  • kwargs – Additional keyword arguments passed to torchmetrics.Metric.

References

[1] Xia & Bouganis. Augmenting Softmax Information for Selective Classification with Out-of-Distribution Data. ACCV, 2022..

[2] Narasimhan et al. Plugin Estimators for Selective Classification with Out-of-Distribution Detection..

See also

SCODCovAtxRisk:

General metric with a configurable risk threshold.