中文
相关论文

相关论文: The Dynamics of Gradient Descent for Overparametri…

200 篇论文

The optimization algorithms are crucial in training physics-informed neural networks (PINNs), as unsuitable methods may lead to poor solutions. Compared to the common gradient descent (GD) algorithm, implicit gradient descent (IGD)…

机器学习 · 计算机科学 2025-08-04 Xianliang Xu , Ting Du , Wang Kong , Bin Shan , Ye Li , Zhongyi Huang

This work is substituted by the paper in arXiv:2011.14066. Stochastic gradient descent is the de facto algorithm for training deep neural networks (DNNs). Despite its popularity, it still requires fine tuning in order to achieve its best…

机器学习 · 统计学 2020-12-02 Vatsal Shah , Anastasios Kyrillidis , Sujay Sanghavi

Deep neural networks achieve stellar generalisation even when they have enough parameters to easily fit all their training data. We study this phenomenon by analysing the dynamics and the performance of over-parameterised two-layer neural…

机器学习 · 统计学 2022-03-28 Sebastian Goldt , Madhu S. Advani , Andrew M. Saxe , Florent Krzakala , Lenka Zdeborová

Most prior work on the convergence of gradient descent (GD) for overparameterized neural networks relies on strong assumptions on the step size (infinitesimal), the hidden-layer width (infinite), or the initialization (large, spectral,…

机器学习 · 计算机科学 2025-05-20 Ziqing Xu , Hancheng Min , Salma Tarmoun , Enrique Mallada , Rene Vidal

In the context of over-parameterization, there is a line of work demonstrating that randomly initialized (stochastic) gradient descent (GD) converges to a globally optimal solution at a linear convergence rate for the quadratic loss…

机器学习 · 计算机科学 2025-06-16 Xianliang Xu , Ting Du , Wang Kong , Bin Shan , Ye Li , Zhongyi Huang

We analyze recurrent neural networks with diagonal hidden-to-hidden weight matrices, trained with gradient descent in the supervised learning setting, and prove that gradient descent can achieve optimality \emph{without} massive…

机器学习 · 计算机科学 2024-10-11 Semih Cayci , Atilla Eryilmaz

There has been a recent surge of interest in understanding the convergence of gradient descent (GD) and stochastic gradient descent (SGD) in overparameterized neural networks. Most previous works assume that the training data is provided a…

机器学习 · 统计学 2021-05-04 Jiaming Xu , Hanjing Zhu

One of the mysteries in the success of neural networks is randomly initialized first order methods like gradient descent can achieve zero training loss even though the objective function is non-convex and non-smooth. This paper demystifies…

机器学习 · 计算机科学 2019-02-06 Simon S. Du , Xiyu Zhai , Barnabas Poczos , Aarti Singh

Deep neural networks are often trained in the over-parametrized regime (i.e. with far more parameters than training examples), and understanding why the training converges to solutions that generalize remains an open problem. Several…

机器学习 · 统计学 2018-03-23 Hartmut Maennel , Olivier Bousquet , Sylvain Gelly

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

Generalization error bounds for deep neural networks trained by stochastic gradient descent (SGD) are derived by combining a dynamical control of an appropriate parameter norm and the Rademacher complexity estimate based on parameter norms.…

机器学习 · 计算机科学 2023-05-30 Mingze Wang , Chao Ma

We study the convergence properties of gradient descent for training deep linear neural networks, i.e., deep matrix factorizations, by extending a previous analysis for the related gradient flow. We show that under suitable conditions on…

机器学习 · 计算机科学 2021-11-25 Gabin Maxime Nguegnang , Holger Rauhut , Ulrich Terstiege

It has been shown that gradient descent can yield the zero training loss in the over-parametrized regime (the width of the neural networks is much larger than the number of data points). In this work, combining the ideas of some existing…

最优化与控制 · 数学 2019-11-05 Lei Li

Two aspects of neural networks that have been extensively studied in the recent literature are their function approximation properties and their training by gradient descent methods. The approximation problem seeks accurate approximations…

机器学习 · 计算机科学 2022-09-20 R. Gentile , G. Welper

The great success neural networks have achieved is inseparable from the application of gradient-descent (GD) algorithms. Based on GD, many variant algorithms have emerged to improve the GD optimization process. The gradient for…

机器学习 · 计算机科学 2023-05-29 Zefan Li , Bingbing Ni , Teng Li , WenJun Zhang , Wen Gao

A main puzzle of deep networks revolves around the absence of overfitting despite large overparametrization and despite the large capacity demonstrated by zero training error on randomly labeled data. In this note, we show that the dynamics…

The simplicity of gradient descent (GD) made it the default method for training ever-deeper and complex neural networks. Both loss functions and architectures are often explicitly tuned to be amenable to this basic local optimization. In…

机器学习 · 计算机科学 2019-04-30 Dmitrii Marin , Meng Tang , Ismail Ben Ayed , Yuri Boykov

Finding parameters in a deep neural network (NN) that fit training data is a nonconvex optimization problem, but a basic first-order optimization method (gradient descent) finds a global optimizer with perfect fit (zero-loss) in many…

机器学习 · 计算机科学 2025-03-07 Zhiyan Ding , Shi Chen , Qin Li , Stephen Wright

Many modern learning tasks involve fitting nonlinear models to data which are trained in an overparameterized regime where the parameters of the model exceed the size of the training dataset. Due to this overparameterization, the training…

机器学习 · 计算机科学 2018-12-27 Samet Oymak , Mahdi Soltanolkotabi

Overparameterization in deep learning typically refers to settings where a trained neural network (NN) has representational capacity to fit the training data in many ways, some of which generalize well, while others do not. In the case of…

机器学习 · 计算机科学 2023-03-24 Edo Cohen-Karlik , Itamar Menuhin-Gruman , Raja Giryes , Nadav Cohen , Amir Globerson