Skip to main navigation Skip to search Skip to main content

Harmless interpolation in regression and classification with structured features

  • College of Computing
  • Michigan State University

Research output: Contribution to journalConference articlepeer-review

2 Citations (Scopus)

Abstract

Overparametrized neural networks tend to perfectly fit noisy training data yet generalize well on test data. Inspired by this empirical observation, recent work has sought to understand this phenomenon of benign overfitting or harmless interpolation in the much simpler linear model. Previous theoretical work critically assumes that either the data features are statistically independent or the input data is high-dimensional; this precludes general nonparametric settings with structured feature maps. In this paper, we present a general and flexible framework for upper bounding regression and classification risk in a reproducing kernel Hilbert space. A key contribution is that our framework describes precise sufficient conditions on the data Gram matrix under which harmless interpolation occurs. Our results recover prior independent-features results (with a much simpler analysis), but they furthermore show that harmless interpolation can occur in more general settings such as features that are a bounded orthonormal system. Furthermore, our results show an asymptotic separation between classification and regression performance in a manner that was previously only shown for Gaussian features.

Original languageEnglish
Pages (from-to)5873-5875
Number of pages3
JournalProceedings of Machine Learning Research
Volume151
Publication statusPublished - 1 Jan 2022
Externally publishedYes
Event25th International Conference on Artificial Intelligence and Statistics, AISTATS 2022 - Virtual, Online, Spain
Duration: 28 Mar 202230 Mar 2022

Fingerprint

Dive into the research topics of 'Harmless interpolation in regression and classification with structured features'. Together they form a unique fingerprint.

Cite this