BetaNLL#

class torch_uncertainty.losses.BetaNLL(beta=0.5, reduction='mean')[source]#

The Beta Negative Log-likelihood loss.

Parameters:
  • beta (float) – Parameter from range [0, 1] controlling relative weighting between data points, where 0 corresponds to high weight on low error points and 1 to an equal weighting.

  • reduction (str, optional) – specifies the reduction to apply to the output:'none' | 'mean' | 'sum'.

References

[1] Seitzer, M., Tavakoli, A., Antic, D., & Martius, G. (2022). On the pitfalls of heteroscedastic uncertainty estimation with probabilistic neural networks.