English
Related papers

Related papers: Interpolation can hurt robust generalization even …

200 papers

We consider correlated \emph{factor} regression models (FRM) and analyze the performance of classical ridge interpolators. Utilizing powerful \emph{Random Duality Theory} (RDT) mathematical engine, we obtain \emph{precise} closed form…

Machine Learning · Statistics 2024-06-14 Mihailo Stojnic

Understanding how overparameterized neural networks generalize despite perfect interpolation of noisy training data is a fundamental question. Mallinar et. al. 2022 noted that neural networks seem to often exhibit ``tempered overfitting'',…

Machine Learning · Computer Science 2024-03-25 Nirmit Joshi , Gal Vardi , Nathan Srebro

The rule of thumb regarding the relationship between the bias-variance tradeoff and model size plays a key role in classical machine learning, but is now well-known to break down in the overparameterized setting as per the double descent…

Machine Learning · Statistics 2026-02-16 Qingyi Hu , Liam Hodgkinson

We study the generalization error of functions that interpolate prescribed data points and are selected by minimizing a weighted norm. Under natural and general conditions, we prove that both the interpolants and their generalization errors…

Numerical Analysis · Mathematics 2021-02-11 Weilin Li

We consider $L^2$-regularized linear (ridge) regression over a finite data sample $X$ with bounded covariance and linear prediction targets $y$ with additive isotropic noise of finite variance. We present an iterative procedure to compute…

Machine Learning · Computer Science 2026-05-28 Jack Timmermans , Sergio A. Alvarez

Benign overfitting, the phenomenon where interpolating models generalize well in the presence of noisy data, was first observed in neural network models trained with gradient descent. To better understand this empirical observation, we…

Machine Learning · Computer Science 2025-07-04 Spencer Frei , Niladri S. Chatterji , Peter L. Bartlett

Deep neural networks (DNNs) typically involve a large number of parameters and are trained to achieve zero or near-zero training error. Despite such interpolation, they often exhibit strong generalization performance on unseen data, a…

Machine Learning · Statistics 2026-01-23 Jingfu Peng , Yuhong Yang

Modern machine learning models often employ a huge number of parameters and are typically optimized to have zero training loss; yet surprisingly, they possess near-optimal prediction performance, contradicting classical learning theory. We…

Machine Learning · Statistics 2021-06-08 Zhu Li , Zhi-Hua Zhou , Arthur Gretton

We study the robust interpolation problem of arbitrary data distributions supported on a bounded space and propose a two-fold law of robustness. Robust interpolation refers to the problem of interpolating $n$ noisy training data points in…

Machine Learning · Computer Science 2023-06-02 Yihan Wu , Heng Huang , Hongyang Zhang

In high dimensional regression, where the number of covariates is of the order of the number of observations, ridge penalization is often used as a remedy against overfitting. Unfortunately, for correlated covariates such regularisation…

Statistics Theory · Mathematics 2023-06-21 Emanuele Massa , Marianne Jonker , Anthony Coolen

Overparameterized models may have many interpolating solutions; implicit regularization refers to the hidden preference of a particular optimization method towards a certain interpolating solution among the many. A by now established line…

Machine Learning · Computer Science 2024-09-18 Hung-Hsu Chou , Holger Rauhut , Rachel Ward

A common strategy to train deep neural networks (DNNs) is to use very large architectures and to train them until they (almost) achieve zero training error. Empirically observed good generalization performance on test data, even in the…

Machine Learning · Statistics 2021-07-26 Nicole Mücke , Ingo Steinwart

We consider the problem of training a model under the presence of label noise. Current approaches identify samples with potentially incorrect labels and reduce their influence on the learning process by either assigning lower weights to…

Machine Learning · Computer Science 2019-06-04 Duc Tam Nguyen , Thi-Phuong-Nhung Ngo , Zhongyu Lou , Michael Klar , Laura Beggel , Thomas Brox

Training classifiers under fairness constraints such as group fairness, regularizes the disparities of predictions between the groups. Nevertheless, even though the constraints are satisfied during training, they might not generalize at…

Machine Learning · Computer Science 2021-03-12 Ching-Yao Chuang , Youssef Mroueh

We consider the overfitting behavior of minimum norm interpolating solutions of Gaussian kernel ridge regression (i.e. kernel ridgeless regression), when the bandwidth or input dimension varies with the sample size. For fixed dimensions, we…

Machine Learning · Computer Science 2024-09-09 Marko Medvedev , Gal Vardi , Nathan Srebro

We show that typical implicit regularization assumptions for deep neural networks (for regression) do not hold for coordinate-MLPs, a family of MLPs that are now ubiquitous in computer vision for representing high-frequency signals. Lack of…

Machine Learning · Computer Science 2022-02-03 Sameera Ramasinghe , Lachlan MacDonald , Simon Lucey

We prove bounds on the population risk of the maximum margin algorithm for two-class linear classification. For linearly separable training data, the maximum margin algorithm has been shown in previous work to be equivalent to a limit of…

Machine Learning · Statistics 2021-06-03 Niladri S. Chatterji , Philip M. Long

Model collapse occurs when generative models degrade after repeatedly training on their own synthetic outputs. We study this effect in overparameterized linear regression in a setting where each iteration mixes fresh real labels with…

Machine Learning · Statistics 2026-02-13 Anvit Garg , Sohom Bhattacharya , Pragya Sur

We consider bounds on the generalization performance of the least-norm linear regressor, in the over-parameterized regime where it can interpolate the data. We describe a sense in which any generalization bound of a type that is commonly…

Machine Learning · Statistics 2021-10-19 Peter L. Bartlett , Philip M. Long

A conventional wisdom in statistical learning is that large models require strong regularization to prevent overfitting. Here we show that this rule can be violated by linear regression in the underdetermined $n\ll p$ situation under…

Statistics Theory · Mathematics 2024-06-06 Dmitry Kobak , Jonathan Lomond , Benoit Sanchez