mixup: Beyond Empirical Risk Minimization

2018 inproceedings zhang2018mixup ○ Not yet verified

Authors

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

Venue

ICLR 2018

Abstract

Introduces mixup, a data augmentation technique that trains on convex combinations of input pairs and their labels. Simple, data-independent, and model-agnostic approach that improves generalization and robustness.

BibTeX

Local Entry

@inproceedings{zhang2018mixup,
  title = {mixup: Beyond Empirical Risk Minimization},
  author = {Hongyi Zhang and Moustapha Cisse and Yann N. Dauphin and David Lopez-Paz},
  year = {2018},
  booktitle = {ICLR 2018},
  url = {https://arxiv.org/abs/1710.09412},
  abstract = {Introduces mixup, a data augmentation technique that trains on convex combinations of input pairs and their labels. Simple, data-independent, and model-agnostic approach that improves generalization and robustness.}
}

From OPENALEX

@inproceedings{zhang2018mixup,
  title = {mixup: Beyond Empirical Risk Minimization},
  author = {Hongyi Zhang and Moustapha Cissé and Yann Dauphin and David Lopez‐Paz},
  year = {2017},
  booktitle = {arXiv (Cornell University)},
  doi = {10.48550/arxiv.1710.09412}
}