English
Related papers

Related papers: Cautious Weight Decay

200 papers

Recently, several works have shown that natural modifications of the classical conditional gradient method (aka Frank-Wolfe algorithm) for constrained convex optimization, provably converge with a linear rate when: i) the feasible set is a…

Optimization and Control · Mathematics 2016-05-23 Dan Garber , Ofer Meshi

We consider distributed optimization under communication constraints for training deep learning models. We propose a new algorithm, whose parameter updates rely on two forces: a regular gradient step, and a corrective direction dictated by…

Machine Learning · Computer Science 2022-04-29 Yunfei Teng , Wenbo Gao , Francois Chalus , Anna Choromanska , Donald Goldfarb , Adrian Weller

Applying weight decay (WD) to matrix layers is standard practice in large-language-model pretraining. Prior work suggests that stochastic gradient noise induces a Brownian-like expansion of the weight matrices W, whose growth is…

Machine Learning · Computer Science 2026-01-09 Maksim Velikanov , Ilyas Chahed , Jingwei Zuo , Dhia Eddine Rhaiem , Younes Belkada , Hakim Hacid

Weakly-supervised salient object detection (WSOD) aims to develop saliency models using image-level annotations. Despite of the success of previous works, explorations on an effective training strategy for the saliency network and accurate…

Computer Vision and Pattern Recognition · Computer Science 2021-09-07 Yongri Piao , Jian Wang , Miao Zhang , Zhengxuan Ma , Huchuan Lu

Weight quantization is an effective technique to compress deep neural networks for their deployment on edge devices with limited resources. Traditional loss-aware quantization methods commonly use the quantized gradient to replace the…

Machine Learning · Computer Science 2024-01-31 Lianbo Ma , Yuee Zhou , Jianlun Ma , Guo Yu , Qing Li

We introduce a novel bias-variance decomposition for a range of strictly convex margin losses, including the logistic loss (minimized by the classic LogitBoost algorithm), as well as the squared margin loss and canonical boosting loss.…

Machine Learning · Statistics 2022-04-27 Danny Wood , Tingting Mu , Gavin Brown

Prediction-correction algorithms are a highly effective class of methods for solving pseudo-convex optimization problems. The descent direction of these algorithms can be viewed as an adjustment to the gradient direction based on the…

Optimization and Control · Mathematics 2025-12-05 Ting Li , Deren Han , Tanxing Wang , Xingju Cai

In this paper we present a novel randomized block coordinate descent method for the minimization of a convex composite objective function. The method uses (approximate) partial second-order (curvature) information, so that the algorithm…

Optimization and Control · Mathematics 2015-05-11 Kimon Fountoulakis , Rachael Tappenden

We introduce a novel framework for learning in neural networks by decomposing each neuron's weight vector into two distinct parts, $W_1$ and $W_2$, thereby modeling contrastive information directly at the neuron level. Traditional gradient…

Machine Learning · Computer Science 2025-03-19 Xi Wang

High dimension low sample size statistical analysis is important in a wide range of applications. In such situations, the highly appealing discrimination method, support vector machine, can be improved to alleviate data piling at the…

Optimization and Control · Mathematics 2017-08-18 Xin Yee Lam , J. S. Marron , Defeng Sun , Kim-Chuan Toh

Randomized coordinate descent (RCD) is a popular optimization algorithm with wide applications in solving various machine learning problems, which motivates a lot of theoretical analysis on its convergence behavior. As a comparison, there…

Machine Learning · Computer Science 2021-08-18 Puyu Wang , Liang Wu , Yunwen Lei

Momentum strategies are an important part of alternative investments and are at the heart of commodity trading advisors (CTAs). These strategies have, however, been found to have difficulties adjusting to rapid changes in market conditions,…

Machine Learning · Statistics 2021-12-21 Kieran Wood , Stephen Roberts , Stefan Zohren

We study a persistent failure mode in multi-objective alignment for large language models (LLMs): training improves performance on only a subset of objectives while causing others to degrade. We formalize this phenomenon as cross-objective…

Computation and Language · Computer Science 2026-05-07 Yining Lu , Meng Jiang

Image segmentation is the foundation of several computer vision tasks, where pixel-wise knowledge is a prerequisite for achieving the desired target. Deep learning has shown promising performance in supervised image segmentation. However,…

Image and Video Processing · Electrical Eng. & Systems 2024-03-19 Boujemaa Guermazi , Riadh Ksantini , Naimul Khan

Consistency Models (CMs) have shown promise for efficient one-step generation. However, most existing CMs rely on manually designed discretization schemes, which can cause repeated adjustments for different noise schedules and datasets. To…

Machine Learning · Computer Science 2025-10-21 Jiayu Bai , Zhanbo Feng , Zhijie Deng , Tianqi Hou , Robert C. Qiu , Zenan Ling

Conventional gradient descent methods compute the gradients for multiple variables through the partial derivative. Treating the coupled variables independently while ignoring the interaction, however, leads to an insufficient optimization…

Machine Learning · Computer Science 2021-06-22 Runqi Wang , Baochang Zhang , Li'an Zhuo , Qixiang Ye , David Doermann

Loss of trainability refers to a phenomenon in continual learning where parameter updates no longer make progress on the optimization objective, so accuracy stalls or degrades as the learning problem changes over time. In this paper, we…

Machine Learning · Computer Science 2025-12-11 Gunbir Singh Baveja , Alex Lewandowski , Mark Schmidt

Prompt tuning, which adapts vision-language models by freezing model parameters and optimizing only the prompt, has proven effective for task-specific adaptations. The core challenge in prompt tuning is improving specialization for a…

Computer Vision and Pattern Recognition · Computer Science 2025-06-10 Dasol Hong , Wooju Lee , Hyun Myung

Sufficient dimension reduction aims for reduction of dimensionality of a regression without loss of information by replacing the original predictor with its lower-dimensional subspace. Partial (sufficient) dimension reduction arises when…

Methodology · Statistics 2019-09-27 Lu Li , Kai Tan , Xuerong Meggie Wen , Zhou Yu

Although ADAM is a very popular algorithm for optimizing the weights of neural networks, it has been recently shown that it can diverge even in simple convex optimization examples. Several variants of ADAM have been proposed to circumvent…

Optimization and Control · Mathematics 2020-09-25 Anas Barakat , Pascal Bianchi
‹ Prev 1 8 9 10 Next ›