English

Foolish Crowds Support Benign Overfitting

Machine Learning 2022-03-21 v5 Machine Learning Statistics Theory Statistics Theory

Abstract

We prove a lower bound on the excess risk of sparse interpolating procedures for linear regression with Gaussian data in the overparameterized regime. We apply this result to obtain a lower bound for basis pursuit (the minimum 1\ell_1-norm interpolant) that implies that its excess risk can converge at an exponentially slower rate than OLS (the minimum 2\ell_2-norm interpolant), even when the ground truth is sparse. Our analysis exposes the benefit of an effect analogous to the "wisdom of the crowd", except here the harm arising from fitting the noise\textit{noise} is ameliorated by spreading it among many directions -- the variance reduction arises from a foolish\textit{foolish} crowd.

Keywords

Cite

@article{arxiv.2110.02914,
  title  = {Foolish Crowds Support Benign Overfitting},
  author = {Niladri S. Chatterji and Philip M. Long},
  journal= {arXiv preprint arXiv:2110.02914},
  year   = {2022}
}
R2 v1 2026-06-24T06:40:41.170Z