Wasserstein Dropout

Despite of its importance for safe machine learning, uncertainty quantification for neural networks is far from being solved. State-of-the-art approaches to estimate neural uncertainties are often hybrid, combining parametric models with explicit or implicit (dropout-based) ensembling. We take another pathway and propose a novel approach to uncertainty quantification for regression tasks, Wasserstein dropout, that is purely non-parametric. Technically, it captures aleatoric uncertainty by means of dropout-based sub-network distributions. This is accomplished by a new objective which minimizes the Wasserstein distance between the label distribution and the model distribution. An extensive empirical analysis shows that Wasserstein dropout outperforms state-of-the-art methods, on vanilla test data as well as under distributional shift in terms of producing more accurate and stable uncertainty estimates.

  • Published in:
    Machine Learning
  • Type:
  • Authors:
    J. Sicking, M. Akila, M. Pintz, T. Wirtz, S. Wrobel, A. Fischer
  • Year:

Citation information

J. Sicking, M. Akila, M. Pintz, T. Wirtz, S. Wrobel, A. Fischer: Wasserstein Dropout, Machine Learning, 2022, https://doi.org/10.1007/s10994-022-06230-8, Sicking.etal.2022,