Wasserstein Distributional Normalization For
         Robust Distributional Certification of Noisy Labeled Data

                       Sung Woo Park 1 Junseok Kwon 1
             Abstract                While there are several methods that can deal with noisy-
    We propose a novel Wasserstein distributional       labeled data, recent methods typically adopt the small-loss
    normalization method that can classify noisy la-      criterion, which helps to construct classification models that
  ...