中文
相关论文

相关论文: An Improved Convergence Analysis of Stochastic Var…

200 篇论文

In decentralized learning, a network of nodes cooperate to minimize an overall objective function that is usually the finite-sum of their local objectives, and incorporates a non-smooth regularization term for the better generalization…

机器学习 · 计算机科学 2022-01-25 Xuanjie Li , Yuedong Xu , Jessie Hui Wang , Xin Wang , John C. S. Lui

In order to model risk aversion in reinforcement learning, an emerging line of research adapts familiar algorithms to optimize coherent risk functionals, a class that includes conditional value-at-risk (CVaR). Because optimizing the…

机器学习 · 计算机科学 2021-03-09 Audrey Huang , Liu Leqi , Zachary C. Lipton , Kamyar Azizzadenesheli

Distributional reinforcement learning (DRL) is a recent reinforcement learning framework whose success has been supported by various empirical studies. It relies on the key idea of replacing the expected return with the return distribution,…

机器学习 · 计算机科学 2020-01-09 Rahul Singh , Keuntaek Lee , Yongxin Chen

Several useful variance-reduced stochastic gradient algorithms, such as SVRG, SAGA, Finito, and SAG, have been proposed to minimize empirical risks with linear convergence properties to the exact minimizer. The existing convergence results…

机器学习 · 计算机科学 2018-02-19 Bicheng Ying , Kun Yuan , Ali H. Sayed

This paper develops the first policy gradient method with global optimality guarantee and complexity analysis for robust reinforcement learning under model mismatch. Robust reinforcement learning is to learn a policy robust to model…

机器学习 · 计算机科学 2022-05-17 Yue Wang , Shaofeng Zou

Since its introduction a decade ago, \emph{relative entropy policy search} (REPS) has demonstrated successful policy learning on a number of simulated and real-world robotic domains, not to mention providing algorithmic components used by…

机器学习 · 计算机科学 2021-03-18 Aldo Pacchiano , Jonathan Lee , Peter Bartlett , Ofir Nachum

The aim of this paper is to deepen the convergence analysis of the scaled gradient projection (SGP) method, proposed by Bonettini et al. in a recent paper for constrained smooth optimization. The main feature of SGP is the presence of a…

数值分析 · 数学 2015-09-10 Silvia Bonettini , Marco Prato

Learning representation from relative similarity comparisons, often called ordinal embedding, gains rising attention in recent years. Most of the existing methods are based on semi-definite programming (\textit{SDP}), which is generally…

机器学习 · 计算机科学 2019-12-03 Ke Ma , Jinshan Zeng , Qianqian Xu , Xiaochun Cao , Wei Liu , Yuan Yao

Deep Q-learning algorithms often suffer from poor gradient estimations with an excessive variance, resulting in unstable training and poor sampling efficiency. Stochastic variance-reduced gradient methods such as SVRG have been applied to…

机器学习 · 计算机科学 2020-07-28 Haonan Jia , Xiao Zhang , Jun Xu , Wei Zeng , Hao Jiang , Xiaohui Yan , Ji-Rong Wen

Variance-reduced stochastic gradient methods have gained popularity in recent times. Several variants exist with different strategies for the storing and sampling of gradients and this work concerns the interactions between these two…

最优化与控制 · 数学 2022-10-19 Martin Morin , Pontus Giselsson

The policy gradient approach is a flexible and powerful reinforcement learning method particularly for problems with continuous actions such as robot control. A common challenge in this scenario is how to reduce the variance of policy…

机器学习 · 计算机科学 2013-01-18 Tingting Zhao , Hirotaka Hachiya , Voot Tangkaratt , Jun Morimoto , Masashi Sugiyama

In the context of finite sums minimization, variance reduction techniques are widely used to improve the performance of state-of-the-art stochastic gradient methods. Their practical impact is clear, as well as their theoretical properties.…

最优化与控制 · 数学 2024-08-07 Cheik Traoré , Vassilis Apidopoulos , Saverio Salzo , Silvia Villa

We consider the problem of offline reinforcement learning (RL) -- a well-motivated setting of RL that aims at policy optimization using only historical data. Despite its wide applicability, theoretical understandings of offline RL, such as…

机器学习 · 计算机科学 2021-02-04 Ming Yin , Yu Bai , Yu-Xiang Wang

Projected policy gradient (PPG) is a basic policy optimization method in reinforcement learning. Given access to exact policy evaluations, previous studies have established the sublinear convergence of PPG for sufficiently small step sizes…

最优化与控制 · 数学 2024-09-19 Jiacai Liu , Wenye Li , Dachao Lin , Ke Wei , Zhihua Zhang

Robotic systems must be able to quickly and robustly make decisions when operating in uncertain and dynamic environments. While Reinforcement Learning (RL) can be used to compute optimal policies with little prior knowledge about the…

机器人学 · 计算机科学 2016-09-13 Yunpeng Pan , Xinyan Yan , Evangelos Theodorou , Byron Boots

We consider a regularized expected reward optimization problem in the non-oblivious setting that covers many existing problems in reinforcement learning (RL). In order to solve such an optimization problem, we apply and analyze the…

机器学习 · 计算机科学 2024-08-21 Ling Liang , Haizhao Yang

As a key component of large language model (LLM) post-training, Reinforcement Learning from Verifiable Rewards (RLVR) has substantially improved reasoning performance. However, existing RLVR algorithms exhibit distinct stability issues:…

计算与语言 · 计算机科学 2026-04-13 Kun Yang , Zikang chen , Yanmeng Wang , Zhigen Li , Ning Cheng , Shaojun Wang , Jing Xiao

Asynchronous reinforcement learning has become increasingly central to scaling LLM post-training, delivering major throughput gains by decoupling rollout generation from policy updates. However, widely used policy-gradient objectives such…

机器学习 · 计算机科学 2026-03-03 Luke J. Huang , Zhuoyang Zhang , Qinghao Hu , Shang Yang , Song Han

In recent years, stochastic variance reduction algorithms have attracted considerable attention for minimizing the average of a large but finite number of loss functions. This paper proposes a novel Riemannian extension of the Euclidean…

机器学习 · 计算机科学 2019-06-03 Hiroyuki Sato , Hiroyuki Kasai , Bamdev Mishra

Reinforcement learning with verifiable rewards (RLVR) has become a practical route to improve large language model reasoning, and Group Relative Policy Optimization (GRPO) is a widely used optimizer in this setting. However, RLVR training…

机器学习 · 计算机科学 2026-05-14 Tue Le , Linh Ngo Van , Trung Le