Jak sobie radzić z tym problemem?

  1. mixup Beyond Empirical Risk Miminization
  2. learning with label noise

Classification in the presence of label ´ noise: A survey - pub, 2013 rok, ale odnosi się ogólnie do ML, a nie szczególnie do Deep Learning

“A survey of label-noise representation learning: Past, present and future” Han,

Wasserstein adversarial regularization (WAR) on label noise” - input perturbation a zarazem leraning with noisy-labels