ML Anthology

Open Problem: How Fast Can a Multiclass Test Set Be Overfit?

Vitaly Feldman, Roy Frostig, Moritz Hardt
COLT 2019 pp. 3185-3189
/colt/2019/feldman2019colt-open/

Abstract

We ask how many measurements of the accuracy on a multiclass benchmark are needed to achieve a given amount of overfitting.

PDF COLT Semantic Scholar

Cite

Text

Feldman et al. "Open Problem: How Fast Can a Multiclass Test Set Be Overfit?." Conference on Learning Theory, 2019.

Markdown

[Feldman et al. "Open Problem: How Fast Can a Multiclass Test Set Be Overfit?." Conference on Learning Theory, 2019.](https://mlanthology.org/colt/2019/feldman2019colt-open/)

BibTeX

@inproceedings{feldman2019colt-open,
  title     = {{Open Problem: How Fast Can a Multiclass Test Set Be Overfit?}},
  author    = {Feldman, Vitaly and Frostig, Roy and Hardt, Moritz},
  booktitle = {Conference on Learning Theory},
  year      = {2019},
  pages     = {3185-3189},
  volume    = {99},
  url       = {https://mlanthology.org/colt/2019/feldman2019colt-open/}
}