中文
相关论文

相关论文: Global Dynamics of Heavy-Tailed SGDs in Nonconvex …

200 篇论文

The empirical success of deep learning is often attributed to SGD's mysterious ability to avoid sharp local minima in the loss landscape, as sharp minima are known to lead to poor generalization. Recently, empirical evidence of heavy-tailed…

机器学习 · 计算机科学 2022-05-12 Xingyu Wang , Sewoong Oh , Chang-Han Rhee

In recent years, various notions of capacity and complexity have been proposed for characterizing the generalization properties of stochastic gradient descent (SGD) in deep learning. Some of the popular notions that correlate well with the…

最优化与控制 · 数学 2021-06-15 Mert Gurbuzbalaban , Umut Şimşekli , Lingjiong Zhu

Stochastic Gradient Descent (SGD) is a cornerstone of large-scale optimization, yet its theoretical behavior under heavy-tailed noise -- common in modern machine learning and reinforcement learning -- remains poorly understood. In this…

最优化与控制 · 数学 2025-08-08 Ilyas Fatkhullin , Florian Hübler , Guanghui Lan

Previous works on stochastic gradient descent (SGD) often focus on its success. In this work, we construct worst-case optimization problems illustrating that, when not in the regimes that the previous works often assume, SGD can exhibit…

机器学习 · 计算机科学 2023-05-30 Liu Ziyin , Botao Li , James B. Simon , Masahito Ueda

It has repeatedly been observed that loss minimization by stochastic gradient descent (SGD) leads to heavy-tailed distributions of neural network parameters. Here, we analyze a continuous diffusion approximation of SGD, called homogenized…

机器学习 · 统计学 2024-02-05 Zhe Jiao , Martin Keller-Ressel

Recent studies have shown that heavy tails can emerge in stochastic optimization and that the heaviness of the tails have links to the generalization error. While these studies have shed light on interesting aspects of the generalization…

机器学习 · 统计学 2023-02-14 Anant Raj , Melih Barsbey , Mert Gürbüzbalaban , Lingjiong Zhu , Umut Şimşekli

In the era of large-scale neural network models, optimization algorithms often struggle with generalization due to an overreliance on training loss. One key insight widely accepted in the machine learning community is the idea that wide…

机器学习 · 计算机科学 2025-09-01 Bodu Gong , Gustavo Enrique Batista , Pierre Lafaye de Micheaux

Non-convex optimization problems are ubiquitous in machine learning, especially in Deep Learning. While such complex problems can often be successfully optimized in practice by using stochastic gradient descent (SGD), theoretical analysis…

机器学习 · 计算机科学 2022-02-21 Harsh Vardhan , Sebastian U. Stich

Recent theoretical studies have shown that heavy-tails can emerge in stochastic optimization due to `multiplicative noise', even under surprisingly simple settings, such as linear regression with Gaussian data. While these studies have…

机器学习 · 统计学 2025-05-06 Mert Gurbuzbalaban , Yuanhan Hu , Umut Simsekli , Kun Yuan , Lingjiong Zhu

Heavy-tail phenomena in stochastic gradient descent (SGD) have been reported in several empirical studies. Experimental evidence in previous works suggests a strong interplay between the heaviness of the tails and generalization behavior of…

机器学习 · 统计学 2023-01-31 Anant Raj , Lingjiong Zhu , Mert Gürbüzbalaban , Umut Şimşekli

In this paper, we examine the time it takes for stochastic gradient descent (SGD) to reach the global minimum of a general, non-convex loss function. We approach this question through the lens of randomly perturbed dynamical systems and…

最优化与控制 · 数学 2025-06-10 Waïss Azizian , Franck Iutzeler , Jérôme Malick , Panayotis Mertikopoulos

Multi-layer neural networks are among the most powerful models in machine learning, yet the fundamental reasons for this success defy mathematical understanding. Learning a neural network requires to optimize a non-convex high-dimensional…

机器学习 · 统计学 2022-06-08 Song Mei , Andrea Montanari , Phan-Minh Nguyen

In machine learning, stochastic gradient descent (SGD) is widely deployed to train models using highly non-convex objectives with equally complex noise models. Unfortunately, SGD theory often makes restrictive assumptions that fail to…

机器学习 · 计算机科学 2022-10-11 Vivak Patel , Shushu Zhang , Bowen Tian

The injection of heavy-tailed noise into the iterates of stochastic gradient descent (SGD) has garnered growing interest in recent years due to its theoretical and empirical benefits for optimization and generalization. However, its…

机器学习 · 统计学 2025-05-13 Umut Şimşekli , Mert Gürbüzbalaban , Sinan Yıldırım , Lingjiong Zhu

Stochastic gradient descent (SGD) is widely used in machine learning. Although being commonly viewed as a fast but not accurate version of gradient descent (GD), it always finds better solutions than GD for modern neural networks. In order…

机器学习 · 计算机科学 2018-08-17 Robert Kleinberg , Yuanzhi Li , Yang Yuan

Understanding the generalization properties of optimization algorithms under heavy-tailed noise has gained growing attention. However, the existing theoretical results mainly focus on stochastic gradient descent (SGD) and the analysis of…

Recent studies have provided both empirical and theoretical evidence illustrating that heavy tails can emerge in stochastic gradient descent (SGD) in various scenarios. Such heavy tails potentially result in iterates with diverging…

最优化与控制 · 数学 2021-02-23 Hongjian Wang , Mert Gürbüzbalaban , Lingjiong Zhu , Umut Şimşekli , Murat A. Erdogdu

Understanding the generalization abilities of modern machine learning algorithms has been a major research topic over the past decades. In recent years, the learning dynamics of Stochastic Gradient Descent (SGD) have been related to…

机器学习 · 统计学 2023-12-04 Benjamin Dupuis , Paul Viallard

Learning with a {\it convex loss} function has been a dominating paradigm for many years. It remains an interesting question how non-convex loss functions help improve the generalization of learning with broad applicability. In this paper,…

机器学习 · 计算机科学 2018-05-22 Yi Xu , Shenghuo Zhu , Sen Yang , Chi Zhang , Rong Jin , Tianbao Yang

The study of tail behaviour of SGD-induced processes has been attracting a lot of interest, due to offering strong guarantees with respect to individual runs of an algorithm. While many works provide high-probability guarantees, quantifying…

机器学习 · 计算机科学 2026-02-06 Aleksandar Armacki , Dragana Bajović , Dušan Jakovetić , Soummya Kar , Ali H. Sayed
‹ 上一页 1 2 3 10 下一页 ›