English
Related papers

Related papers: Rethinking the Global Convergence of Softmax Polic…

200 papers

Entropy regularization is an important idea in reinforcement learning, with great success in recent algorithms like Soft Q Network (SQN) and Soft Actor-Critic (SAC1). In this work, we extend this idea into the on-policy realm. We propose…

Machine Learning · Computer Science 2020-10-19 Jingbin Liu , Xinyang Gu , Shuai Liu

We study problem-dependent rates, i.e., generalization errors that scale near-optimally with the variance, the effective loss, or the gradient norms evaluated at the "best hypothesis." We introduce a principled framework dubbed "uniform…

Machine Learning · Statistics 2020-12-25 Yunbei Xu , Assaf Zeevi

In this paper, we establish the global optimality and convergence rate of an off-policy actor critic algorithm in the tabular setting without using density ratio to correct the discrepancy between the state distribution of the behavior…

Machine Learning · Computer Science 2025-02-07 Shangtong Zhang , Remi Tachet , Romain Laroche

Stochastic gradient methods are among the most important algorithms in training machine learning problems. While classical assumptions such as strong convexity allow a simple analysis they are rarely satisfied in applications. In recent…

Machine Learning · Computer Science 2025-03-18 Simon Weissmann , Sara Klein , Waïss Azizian , Leif Döring

We study the global convergence of policy gradient for infinite-horizon, continuous state and action space, and entropy-regularized Markov decision processes (MDPs). We consider a softmax policy with (one-hidden layer) neural network…

Optimization and Control · Mathematics 2022-06-17 Bekzhan Kerimkulov , James-Michael Leahy , David Šiška , Lukasz Szpruch

We introduce the receding-horizon policy gradient (RHPG) algorithm, the first PG algorithm with provable global convergence in learning the optimal linear estimator designs, i.e., the Kalman filter (KF). Notably, the RHPG algorithm does not…

Optimization and Control · Mathematics 2023-09-12 Xiangyuan Zhang , Saviz Mowlavi , Mouhacine Benosman , Tamer Başar

Reinforcement learning (RL) has gained attention for aligning large language models (LLMs) via reinforcement learning from human feedback (RLHF). The actor-only variants of Proximal Policy Optimization (PPO) are widely applied for their…

Optimization and Control · Mathematics 2025-12-19 Yin Liu , Qiming Dai , Junyu Zhang , Zaiwen Wen

Asynchronous stochastic approximations (SAs) are an important class of model-free algorithms, tools and techniques that are popular in multi-agent and distributed control scenarios. To counter Bellman's curse of dimensionality, such…

Optimization and Control · Mathematics 2019-05-03 Arunselvan Ramaswamy , Shalabh Bhatnagar , Daniel E. Quevedo

We analyze the convergence rate of the unregularized natural policy gradient algorithm with log-linear policy parametrizations in infinite-horizon discounted Markov decision processes. In the deterministic case, when the Q-value is known…

Machine Learning · Computer Science 2023-03-15 Carlo Alfano , Patrick Rebeschini

ReParameterization (RP) Policy Gradient Methods (PGMs) have been widely adopted for continuous control tasks in robotics and computer graphics. However, recent studies have revealed that, when applied to long-term reinforcement learning…

Machine Learning · Computer Science 2023-11-01 Shenao Zhang , Boyi Liu , Zhaoran Wang , Tuo Zhao

Direct policy gradient methods for reinforcement learning are a successful approach for a variety of reasons: they are model free, they directly optimize the performance metric of interest, and they allow for richly parameterized policies.…

Machine Learning · Computer Science 2020-08-14 Alekh Agarwal , Mikael Henaff , Sham Kakade , Wen Sun

Policy gradient is a generic and flexible reinforcement learning approach that generally enjoys simplicity in analysis, implementation, and deployment. In the last few decades, this approach has been extensively advanced for fully…

Machine Learning · Computer Science 2020-05-26 Kamyar Azizzadenesheli , Yisong Yue , Animashree Anandkumar

Efficient exploration is a central problem in reinforcement learning and is often formalized as maximizing the entropy of the state-action occupancy measure. While unconstrained maximum-entropy exploration is relatively well understood,…

Machine Learning · Computer Science 2026-05-01 Florian Wolf , Ilyas Fatkhullin , Niao He

Policy gradient methods with actor-critic schemes demonstrate tremendous empirical successes, especially when the actors and critics are parameterized by neural networks. However, it remains less clear whether such "neural" policy gradient…

Machine Learning · Computer Science 2020-06-23 Lingxiao Wang , Qi Cai , Zhuoran Yang , Zhaoran Wang

Maximum entropy deep reinforcement learning (RL) methods have been demonstrated on a range of challenging continuous tasks. However, existing methods either suffer from severe instability when training on large off-policy data or cannot…

Machine Learning · Computer Science 2019-09-10 Wenjie Shi , Shiji Song , Cheng Wu

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…

Machine Learning · Computer Science 2021-03-09 Audrey Huang , Liu Leqi , Zachary C. Lipton , Kamyar Azizzadenesheli

Stabilizing a dynamical system is a fundamental problem that serves as a cornerstone for many complex tasks in the field of control systems. The problem becomes challenging when the system model is unknown. Among the Reinforcement Learning…

Systems and Control · Electrical Eng. & Systems 2026-01-30 Ankang Zhang , Ming Chi , Xiaoling Wang , Lintao Ye

Policy optimization, which finds the desired policy by maximizing value functions via optimization techniques, lies at the heart of reinforcement learning (RL). In addition to value maximization, other practical considerations arise as…

Machine Learning · Computer Science 2023-01-12 Wenhao Zhan , Shicong Cen , Baihe Huang , Yuxin Chen , Jason D. Lee , Yuejie Chi

This paper proposes a novel termination criterion, termed the advantage gap function, for finite state and action Markov decision processes (MDP) and reinforcement learning (RL). By incorporating this advantage gap function into the design…

Machine Learning · Computer Science 2026-03-24 Caleb Ju , Guanghui Lan

Many popular policy gradient methods for reinforcement learning follow a biased approximation of the policy gradient known as the discounted approximation. While it has been shown that the discounted approximation of the policy gradient is…

Machine Learning · Computer Science 2023-01-10 Chris Nota