中文
相关论文

相关论文: Learning Curves and Benign Overfitting of Spectral…

200 篇论文

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…

机器学习 · 统计学 2021-06-08 Zhu Li , Zhi-Hua Zhou , Arthur Gretton

The success of over-parameterized neural networks trained to near-zero training error has caused great interest in the phenomenon of benign overfitting, where estimators are statistically consistent even though they interpolate noisy…

机器学习 · 统计学 2024-11-07 Moritz Haas , David Holzmüller , Ulrike von Luxburg , Ingo Steinwart

The widely observed 'benign overfitting phenomenon' in the neural network literature raises the challenge to the 'bias-variance trade-off' doctrine in the statistical learning theory. Since the generalization ability of the 'lazy trained'…

机器学习 · 计算机科学 2023-09-26 Yicheng Li , Haobo Zhang , Qian Lin

The generalization ability of kernel interpolation in large dimensions (i.e., $n \asymp d^{\gamma}$ for some $\gamma>0$) might be one of the most interesting problems in the recent renaissance of kernel regression, since it may help us…

机器学习 · 计算机科学 2024-04-22 Haobo Zhang , Weihao Lu , Qian Lin

The remarkable practical success of deep learning has revealed some major surprises from a theoretical perspective. In particular, simple gradient methods easily find near-optimal solutions to non-convex optimization problems, and despite…

统计理论 · 数学 2021-03-17 Peter L. Bartlett , Andrea Montanari , Alexander Rakhlin

The practical success of overparameterized neural networks has motivated the recent scientific study of interpolating methods, which perfectly fit their training data. Certain interpolating methods, including neural networks, can fit noisy…

机器学习 · 计算机科学 2024-07-17 Neil Mallinar , James B. Simon , Amirhesam Abedsoltan , Parthe Pandit , Mikhail Belkin , Preetum Nakkiran

The phenomenon of benign overfitting is one of the key mysteries uncovered by deep learning methodology: deep neural networks seem to predict well, even with a perfect fit to noisy training data. Motivated by this phenomenon, we consider…

机器学习 · 统计学 2022-06-08 Peter L. Bartlett , Philip M. Long , Gábor Lugosi , Alexander Tsigler

Benign overfitting is a phenomenon in machine learning where a model perfectly fits (interpolates) the training data, including noisy examples, yet still generalizes well to unseen data. Understanding this phenomenon has attracted…

机器学习 · 计算机科学 2025-05-20 Junhyung Park , Patrick Bloebaum , Shiva Prasad Kasiviswanathan

The literature on "benign overfitting" in overparameterized models has been mostly restricted to regression or binary classification; however, modern machine learning operates in the multiclass setting. Motivated by this discrepancy, we…

机器学习 · 统计学 2023-07-13 Ke Wang , Vidya Muthukumar , Christos Thrampoulidis

Recent advances in machine learning have inspired a surge of research into reconstructing specific quantities of interest from measurements that comply with certain physical laws. These efforts focus on inverse problems that are governed by…

机器学习 · 统计学 2025-04-23 Honam Wong , Wendao Wu , Fanghui Liu , Yiping Lu

Generalization performance of classifiers in deep learning has recently become a subject of intense study. Deep models, typically over-parametrized, tend to fit the training data exactly. Despite this "overfitting", they perform well on…

机器学习 · 统计学 2018-06-18 Mikhail Belkin , Siyuan Ma , Soumik Mandal

In this chapter we provide a theoretically founded investigation of state-of-the-art learning approaches for inverse problems from the point of view of spectral reconstruction operators. We give an extended definition of regularization…

数值分析 · 数学 2024-06-05 Martin Burger , Samira Kabri

``Benign overfitting'', the ability of certain algorithms to interpolate noisy training data and yet perform well out-of-sample, has been a topic of considerable recent interest. We show, using a fixed design setup, that an important class…

机器学习 · 计算机科学 2023-04-14 Daniel Beaglehole , Mikhail Belkin , Parthe Pandit

Motivated by the studies of neural networks (e.g.,the neural tangent kernel theory), we perform a study on the large-dimensional behavior of kernel ridge regression (KRR) where the sample size $n \asymp d^{\gamma}$ for some $\gamma > 0$.…

机器学习 · 计算机科学 2024-01-03 Haobo Zhang , Yicheng Li , Weihao Lu , Qian Lin

Deep learning methods operate in regimes that defy the traditional statistical mindset. Neural network architectures often contain more parameters than training samples, and are so rich that they can interpolate the observed labels, even if…

统计理论 · 数学 2020-12-14 Song Mei , Andrea Montanari

We perform a study on kernel regression for large-dimensional data (where the sample size $n$ is polynomially depending on the dimension $d$ of the samples, i.e., $n\asymp d^{\gamma}$ for some $\gamma >0$ ). We first build a general tool to…

机器学习 · 统计学 2024-07-01 Weihao Lu , Haobo Zhang , Yicheng Li , Manyun Xu , Qian Lin

As modern machine learning models continue to advance the computational frontier, it has become increasingly important to develop precise estimates for expected performance improvements under different model and data scaling regimes.…

机器学习 · 计算机科学 2023-06-13 Lechao Xiao , Hong Hu , Theodor Misiakiewicz , Yue M. Lu , Jeffrey Pennington

Generalization beyond a training dataset is a main goal of machine learning, but theoretical understanding of generalization remains an open problem for many models. The need for a new theory is exacerbated by recent observations in deep…

机器学习 · 统计学 2022-02-08 Abdulkadir Canatar , Blake Bordelon , Cengiz Pehlevan

Modern deep learning models with great expressive power can be trained to overfit the training data but still generalize well. This phenomenon is referred to as \textit{benign overfitting}. Recently, a few studies have attempted to…

机器学习 · 计算机科学 2023-11-07 Yiwen Kou , Zixiang Chen , Yuanzhou Chen , Quanquan Gu

Modern deep learning models employ considerably more parameters than required to fit the training data. Whereas conventional statistical wisdom suggests such models should drastically overfit, in practice these models generalize remarkably…

机器学习 · 统计学 2020-08-18 Ben Adlam , Jeffrey Pennington
‹ 上一页 1 2 3 10 下一页 ›