mixup: Beyond Empirical Risk Minimization

Hongyi Zhang, Moustapha Cisse, Yann N. Dauphin, David Lopez-Paz

Feb 15, 2018 (modified: Feb 23, 2018) ICLR 2018 Conference Blind Submission readers: everyone Show Bibtex
  • Abstract: Large deep neural networks are powerful, but exhibit undesirable behaviors such as memorization and sensitivity to adversarial examples. In this work, we propose mixup, a simple learning principle to alleviate these issues. In essence, mixup trains a neural network on convex combinations of pairs of examples and their labels. By doing so, mixup regularizes the neural network to favor simple linear behavior in-between training examples. Our experiments on the ImageNet-2012, CIFAR-10, CIFAR-100, Google commands and UCI datasets show that mixup improves the generalization of state-of-the-art neural network architectures. We also find that mixup reduces the memorization of corrupt labels, increases the robustness to adversarial examples, and stabilizes the training of generative adversarial networks.
  • TL;DR: Training on convex combinations between random training examples and their labels improves generalization in deep neural networks
  • Keywords: empirical risk minimization, vicinal risk minimization, generalization, data augmentation, image classification, generative adversarial networks, adversarial examples, random labels