中文
相关论文

相关论文: Double-descent curves in neural networks: a new pe…

200 篇论文

Various Graph Neural Networks (GNNs) have been successful in analyzing data in non-Euclidean spaces, however, they have limitations such as oversmoothing, i.e., information becomes excessively averaged as the number of hidden layers…

机器学习 · 计算机科学 2024-01-23 Jaeyoon Sim , Sooyeon Jeon , InJun Choi , Guorong Wu , Won Hwa Kim

We study the training and generalization of deep neural networks (DNNs) in the over-parameterized regime, where the network width (i.e., number of hidden nodes per layer) is much larger than the number of training data points. We show that,…

机器学习 · 计算机科学 2019-11-13 Yuan Cao , Quanquan Gu

Kernels representing limiting cases of neural network architectures have recently gained popularity. However, the application and performance of these new kernels compared to existing options, such as the Matern kernel, is not well studied.…

机器学习 · 统计学 2024-10-14 Amanda Muyskens , Benjamin W. Priest , Imene R. Goumiri , Michael D. Schneider

Neural tangent kernels (NTKs) provide a theoretical regime to analyze the learning and generalization behavior of over-parametrized neural networks. For a supervised learning task, the association between the eigenvectors of the NTK kernel…

机器学习 · 计算机科学 2023-10-18 Shervin Khalafi , Saurabh Sihag , Alejandro Ribeiro

Graph convolutional neural networks~(GCNs) have recently demonstrated promising results on graph-based semi-supervised classification, but little work has been done to explore their theoretical properties. Recently, several deep neural…

机器学习 · 计算机科学 2020-02-28 Jilin Hu , Jianbing Shen , Bin Yang , Ling Shao

Recent work has shown that training wide neural networks with gradient descent is formally equivalent to computing the mean of the posterior distribution in a Gaussian Process (GP) with the Neural Tangent Kernel (NTK) as the prior…

机器学习 · 计算机科学 2024-09-11 Sergio Calvo-Ordoñez , Konstantina Palla , Kamil Ciosek

We consider the dynamics of gradient descent (GD) in overparameterized single hidden layer neural networks with a squared loss function. Recently, it has been shown that, under some conditions, the parameter values obtained using GD achieve…

机器学习 · 计算机科学 2021-05-17 Siddhartha Satpathi , R Srikant

A long-standing goal in deep learning has been to characterize the learning behavior of black-box models in a more interpretable manner. For graph neural networks (GNNs), considerable advances have been made in formalizing what functions…

机器学习 · 计算机科学 2024-06-19 Chenxiao Yang , Qitian Wu , David Wipf , Ruoyu Sun , Junchi Yan

Classical statistical learning theory predicts that overparameterized models should exhibit severe overfitting, yet modern deep neural networks with far more parameters than training samples consistently generalize well. This contradiction…

机器学习 · 计算机科学 2026-04-10 Zeran Johannsen

We study overparameterization in generative adversarial networks (GANs) that can interpolate the training data. We show that overparameterization can improve generalization performance and accelerate the training process. We study the…

机器学习 · 计算机科学 2024-05-02 Lorenzo Luzi , Yehuda Dar , Richard Baraniuk

Several recent trends in machine learning theory and practice, from the design of state-of-the-art Gaussian Process to the convergence analysis of deep neural nets (DNNs) under stochastic gradient descent (SGD), have found it fruitful to…

神经与进化计算 · 计算机科学 2020-04-07 Greg Yang

Recent work has established the equivalence between deep neural networks and Gaussian processes (GPs), resulting in so-called neural network Gaussian processes (NNGPs). The behaviour of these models depends on the initialisation of the…

机器学习 · 计算机科学 2020-07-15 Arnu Pretorius , Herman Kamper , Steve Kroon

We explore the link between deep ensembles and Gaussian processes (GPs) through the lens of the Neural Tangent Kernel (NTK): a recent development in understanding the training dynamics of wide neural networks (NNs). Previous work has shown…

机器学习 · 统计学 2020-10-27 Bobby He , Balaji Lakshminarayanan , Yee Whye Teh

Recent advances in Deep Gaussian Processes (DGPs) show the potential to have more expressive representation than that of traditional Gaussian Processes (GPs). However, there exists a pathology of deep Gaussian processes that their learning…

机器学习 · 计算机科学 2020-12-22 Anh Tong , Jaesik Choi

Graph neural networks (GNNs) have emerged as powerful tools for processing relational data in applications. However, GNNs suffer from the problem of oversmoothing, the property that the features of all nodes exponentially converge to the…

机器学习 · 统计学 2025-05-22 Bastian Epping , Alexandre René , Moritz Helias , Michael T. Schaub

This paper aims to discuss the impact of random initialization of neural networks in the neural tangent kernel (NTK) theory, which is ignored by most recent works in the NTK theory. It is well known that as the network's width tends to…

机器学习 · 统计学 2024-10-10 Guhan Chen , Yicheng Li , Qian Lin

Extensive empirical evidence reveals that, for a wide range of different learning methods and datasets, the risk curve exhibits a double-descent (DD) trend as a function of the model size. In a recent paper [Zeyu,Kammoun,Thrampoulidis,2019]…

机器学习 · 统计学 2020-02-03 Ganesh Kini , Christos Thrampoulidis

We study the transfer learning process between two linear regression problems. An important and timely special case is when the regressors are overparameterized and perfectly interpolate their training data. We examine a parameter transfer…

机器学习 · 计算机科学 2022-09-29 Yehuda Dar , Richard G. Baraniuk

This study demonstrates that double descent can be mitigated by adding a dropout layer adjacent to the fully connected linear layer. The unexpected double-descent phenomenon garnered substantial attention in recent years, resulting in…

机器学习 · 计算机科学 2025-08-08 Tian-Le Yang , Joe Suzuki

Recent advances in machine learning have been achieved by using overparametrized models trained until near interpolation of the training data. It was shown, e.g., through the double descent phenomenon, that the number of parameters is a…

机器学习 · 统计学 2024-03-14 Hong Hu , Yue M. Lu , Theodor Misiakiewicz