Harmless interpolation in regression and classification with structured features

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.

References in corpus (9)

• Fast learning rates for plug-in classifiers
• Benign overfitting in ridge regression
• Understanding Double Descent Requires a Fine-Grained Bias-Variance Decomposition
• A Precise Performance Analysis of Learning with Random Features
• Universality Laws for High-Dimensional Learning with Random Features
• What causes the test error? Going beyond bias-variance via ANOVA
• Generalization error of random features and kernel methods: hypercontractivity and kernel matrix concentration
• Risk Bounds for Over-parameterized Maximum Margin Classification on Sub-Gaussian Mixtures
• Towards an Understanding of Benign Overfitting in Neural Networks