Gradient-Matching Coresets for Continual Learning

Abstract

We devise a coreset selection method based on the idea of gradient matching: the gradients induced by the coreset should match, as closely as possible, those induced by the original training dataset. We evaluate the method in the context of continual learning, where it can be used to curate a rehearsal memory. Our method performs strong competitors such as reservoir sampling across a range of memory sizes.

PDF NeurIPSW OpenReview Semantic Scholar

Cite

Text

Balles et al. "Gradient-Matching Coresets for Continual Learning." NeurIPS 2021 Workshops: DistShift, 2021.

Markdown

[Balles et al. "Gradient-Matching Coresets for Continual Learning." NeurIPS 2021 Workshops: DistShift, 2021.](https://mlanthology.org/neuripsw/2021/balles2021neuripsw-gradientmatching/)

BibTeX

@inproceedings{balles2021neuripsw-gradientmatching,
  title     = {{Gradient-Matching Coresets for Continual Learning}},
  author    = {Balles, Lukas and Zappella, Giovanni and Archambeau, Cedric},
  booktitle = {NeurIPS 2021 Workshops: DistShift},
  year      = {2021},
  url       = {https://mlanthology.org/neuripsw/2021/balles2021neuripsw-gradientmatching/}
}