中文
相关论文

相关论文: On the linearity of large non-linear models: when …

200 篇论文

Linear networks provide valuable insights into the workings of neural networks in general. This paper identifies conditions under which the gradient flow provably trains a linear network, in spite of the non-strict saddle points present in…

最优化与控制 · 数学 2020-06-30 Armin Eftekhari

Neural networks has recently attracted much interest as useful representations of quantum many body ground states, which might help address the infamous sign problem. Most attention was directed at their representability properties, while…

量子物理 · 物理学 2024-11-07 Harel Kol-Namer , Moshe Goldstein

Past decades have witnessed a great interest in the distinction and connection between neural network learning and kernel learning. Recent advancements have made theoretical progress in connecting infinite-wide neural networks and Gaussian…

机器学习 · 计算机科学 2025-11-07 Shao-Qun Zhang , Zong-Yi Chen , Yong-Ming Tian , Xun Lu

Explaining generalizations and preventing over-confident predictions are central goals of studies on the loss landscape of neural networks. Flatness, defined as loss invariability on perturbations of a pre-trained solution, is widely…

机器学习 · 计算机科学 2022-10-03 SungYub Kim , Sihwan Park , Kyungsu Kim , Eunho Yang

It is well understood that neural networks with carefully hand-picked weights provide powerful function approximation and that they can be successfully trained in over-parametrized regimes. Since over-parametrization ensures zero training…

机器学习 · 计算机科学 2024-05-21 G. Welper

We analyze the generalization properties of two-layer neural networks in the neural tangent kernel (NTK) regime, trained with gradient descent (GD). For early stopped GD we derive fast rates of convergence that are known to be minimax…

机器学习 · 统计学 2023-09-18 Mike Nguyen , Nicole Mücke

In this article a surprising result is demonstrated using the neural tangent kernel. This kernel is defined as the inner product of the vector of the gradient of an underlying model evaluated at training points. This kernel is used to…

人工智能 · 计算机科学 2021-04-14 Matt Calder

Graph neural networks (GNNs) achieve remarkable performance in graph machine learning tasks but can be hard to train on large-graph data, where their learning dynamics are not well understood. We investigate the training dynamics of…

机器学习 · 计算机科学 2023-06-02 Sanjukta Krishnagopal , Luana Ruiz

Existing analyses of neural network training often operate under the unrealistic assumption of an extremely small learning rate. This lies in stark contrast to practical wisdom and empirical studies, such as the work of J. Cohen et al.…

机器学习 · 计算机科学 2023-10-20 Kwangjun Ahn , Sébastien Bubeck , Sinho Chewi , Yin Tat Lee , Felipe Suarez , Yi Zhang

We perform a careful, thorough, and large scale empirical study of the correspondence between wide neural networks and kernel methods. By doing so, we resolve a variety of open questions related to the study of infinitely wide neural…

Recently, neural networks utilizing periodic activation functions have been proven to demonstrate superior performance in vision tasks compared to traditional ReLU-activated networks. However, there is still a limited understanding of the…

机器学习 · 计算机科学 2024-02-08 Hemanth Saratchandran , Shin-Fang Chng , Simon Lucey

Despite the widespread adoption of neural networks, their training dynamics remain poorly understood. We show experimentally that as the size of the dataset increases, a point forms where the magnitude of the gradient of the loss becomes…

机器学习 · 计算机科学 2024-07-23 Mark Lowell

Are neural networks biased toward simple functions? Does depth always help learn more complex features? Is training the last layer of a network as good as training all layers? How to set the range for learning rate tuning? These questions…

机器学习 · 计算机科学 2020-04-10 Greg Yang , Hadi Salman

Recently, quantum neural networks or quantum-classical neural networks (qcNN) have been actively studied, as a possible alternative to the conventional classical neural network (cNN), but their practical and theoretically-guaranteed…

量子物理 · 物理学 2023-12-12 Kouhei Nakaji , Hiroyuki Tezuka , Naoki Yamamoto

Understanding how deep neural networks learn remains a fundamental challenge in modern machine learning. A growing body of evidence suggests that training dynamics undergo a distinct phase transition, yet our understanding of this…

机器学习 · 计算机科学 2025-05-21 Zhanpeng Zhou , Yongyi Yang , Mahito Sugiyama , Junchi Yan

The study of Neural Tangent Kernels (NTKs) in deep learning has drawn increasing attention in recent years. NTKs typically actively change during training and are related to feature learning. In parallel, recent work on Gradient Descent…

机器学习 · 计算机科学 2025-07-18 Kaiqi Jiang , Jeremy Cohen , Yuanzhi Li

We perform a study on the generalization ability of the wide two-layer ReLU neural network on $\mathbb{R}$. We first establish some spectral properties of the neural tangent kernel (NTK): $a)$ $K_{d}$, the NTK defined on $\mathbb{R}^{d}$,…

机器学习 · 统计学 2023-02-14 Jianfa Lai , Manyun Xu , Rui Chen , Qian Lin

Deep neural networks have shown many fruitful applications in this decade. A network can get the generalized function through training with a finite dataset. The degree of generalization is a realization of the proximity scale in the data…

无序系统与神经网络 · 物理学 2023-09-15 Kenichi Nakazato

Data with low-dimensional nonlinear structure are ubiquitous in engineering and scientific problems. We study a model problem with such structure -- a binary classification task that uses a deep fully-connected neural network to classify…

机器学习 · 统计学 2021-11-01 Tingran Wang , Sam Buchanan , Dar Gilboa , John Wright

In multi-objective optimization, multiple loss terms are weighted and added together to form a single objective. These weights are chosen to properly balance the competing losses according to some meta-goal. For example, in physics-informed…

数值分析 · 数学 2025-11-20 Max Hirsch , Federico Pichi