English
Related papers

Related papers: R-GTD: A Geometric Analysis of Gradient Temporal-D…

200 papers

Graph continual learning (GCL) aims to learn from a continuous sequence of graph-based tasks. Regularization methods are vital for preventing catastrophic forgetting in GCL, particularly in the challenging replay-free, class-incremental…

Machine Learning · Computer Science 2025-09-17 Jie Yin , Ke Sun , Han Wu

Recent advances in rule-based reinforcement learning (RL) have significantly improved the reasoning capability of language models (LMs) with rule-based rewards. However, existing RL methods -- such as GRPO, REINFORCE++, and RLOO -- often…

Machine Learning · Computer Science 2025-05-20 Zongkai Liu , Fanqing Meng , Lingxiao Du , Zhixiang Zhou , Chao Yu , Wenqi Shao , Qiaosheng Zhang

We propose graph-dependent implicit regularisation strategies for distributed stochastic subgradient descent (Distributed SGD) for convex problems in multi-agent learning. Under the standard assumptions of convexity, Lipschitz continuity,…

Machine Learning · Computer Science 2018-09-20 Dominic Richards , Patrick Rebeschini

In most practical applications such as recommendation systems, display advertising, and so forth, the collected data often contains missing values and those missing values are generally missing-not-at-random, which deteriorates the…

Machine Learning · Computer Science 2024-05-27 Mingming Ha , Xuewen Tao , Wenfang Lin , Qionxu Ma , Wujiang Xu , Linxun Chen

Optimization problem, which is aimed at finding the global minimal value of a given cost function, is one of the central problem in science and engineering. Various numerical methods have been proposed to solve this problem, among which the…

Optimization and Control · Mathematics 2022-10-07 Shaojun Dong , Fengyu Le , Meng Zhang , Si-Jing Tao , Chao Wang , Yong-Jian Han , Guo-Ping Guo

Value function approximation is a crucial module for policy evaluation in reinforcement learning when the state space is large or continuous. The present paper takes a generative perspective on policy evaluation via temporal-difference (TD)…

Machine Learning · Statistics 2021-12-03 Qin Lu , Georgios B. Giannakis

We analyze inexact Riemannian gradient descent (RGD) where Riemannian gradients and retractions are inexactly (and cheaply) computed. Our focus is on understanding when inexact RGD converges and what is the complexity in the general…

Optimization and Control · Mathematics 2024-05-10 Yuchen Li , Laura Balzano , Deanna Needell , Hanbaek Lyu

We study a type of Riemannian gradient descent (RGD) algorithm, designed through Riemannian preconditioning, for optimization on $\mathcal{M}_k^{m\times n}$ -- the set of $m\times n$ real matrices with a fixed rank $k$. Our analysis is…

Optimization and Control · Mathematics 2024-08-15 Shuyu Dong , Bin Gao , Wen Huang , Kyle A. Gallivan

Off-policy learning enables a reinforcement learning (RL) agent to reason counterfactually about policies that are not executed and is one of the most important ideas in RL. It, however, can lead to instability when combined with function…

Machine Learning · Computer Science 2025-03-03 Xiaochi Qian , Shangtong Zhang

This paper considers the problem of solving constrained reinforcement learning (RL) problems with anytime guarantees, meaning that the algorithmic solution must yield a constraint-satisfying policy at every iteration of its evolution. Our…

Systems and Control · Electrical Eng. & Systems 2025-10-03 Pol Mestres , Arnau Marzabal , Jorge Cortés

We build a theoretical framework for designing and understanding practical meta-learning methods that integrates sophisticated formalizations of task-similarity with the extensive literature on online convex optimization and sequential…

Machine Learning · Computer Science 2019-12-10 Mikhail Khodak , Maria-Florina Balcan , Ameet Talwalkar

In this paper we study the problem of convergence and generalization error bound of stochastic momentum for deep learning from the perspective of regularization. To do so, we first interpret momentum as solving an $\ell_2$-regularized…

Machine Learning · Computer Science 2019-06-04 Ziming Zhang , Wenju Xu , Alan Sullivan

Reward maximization, safe exploration, and intrinsic motivation are often studied as separate objectives in reinforcement learning (RL). We present a unified geometric framework, that views these goals as instances of a single optimization…

Machine Learning · Computer Science 2025-09-03 Nikola Milosevic , Nico Scherf

Excessive computational cost for learning large data and streaming data can be alleviated by using stochastic algorithms, such as stochastic gradient descent and its variants. Recent advances improve stochastic algorithms on convergence…

Machine Learning · Statistics 2019-09-24 Shih-Kang Chao , Guang Cheng

Targeted data selection has emerged as a crucial paradigm for efficient instruction tuning, aiming to identify a small yet influential subset of training examples for a specific target task. In practice, influence is often measured through…

Machine Learning · Computer Science 2026-05-19 Guanghui Min , Tianhao Huang , Ke Wan , Chen Chen

In reinforcement learning (RL) , one of the key components is policy evaluation, which aims to estimate the value function (i.e., expected long-term accumulated reward) of a policy. With a good policy evaluation method, the RL algorithms…

Machine Learning · Computer Science 2018-09-25 Yue Wang , Wei Chen , Yuting Liu , Zhi-Ming Ma , Tie-Yan Liu

Classical optimisation theory guarantees monotonic objective decrease for gradient descent (GD) when employed in a small step size, or ``stable", regime. In contrast, gradient descent on neural networks is frequently performed in a large…

Machine Learning · Computer Science 2025-10-21 Lachlan Ewen MacDonald , Hancheng Min , Leandro Palma , Salma Tarmoun , Ziqing Xu , René Vidal

Gradient descent or its variants are popular in training neural networks. However, in deep Q-learning with neural network approximation, a type of reinforcement learning, gradient descent (also known as Residual Gradient (RG)) is barely…

Machine Learning · Computer Science 2022-11-15 Shuyu Yin , Tao Luo , Peilin Liu , Zhi-Qin John Xu

Entropy regularized algorithms such as Soft Q-learning and Soft Actor-Critic, recently showed state-of-the-art performance on a number of challenging reinforcement learning (RL) tasks. The regularized formulation modifies the standard RL…

Machine Learning · Statistics 2019-10-15 Elena Smirnova , Elvis Dohmatob

A major obstacle to achieving global convergence in distributed and federated learning is the misalignment of gradients across clients, or mini-batches due to heterogeneity and stochasticity of the distributed data. In this work, we show…

Machine Learning · Computer Science 2021-12-14 Yatin Dandi , Luis Barba , Martin Jaggi