paper-with-me

홈 › Papers

State-Action Joint Regularized Implicit Policy for Offline Reinforcement Learning

2021-09-29 · Shentao Yang, Zhendong Wang, Huangjie Zheng, Mingyuan Zhou

Offline reinforcement learning enables learning from a fixed dataset, without further interactions with the environment. The lack of environmental interactions makes the policy training vulnerable to state-action pairs far from the training dataset and prone to missing rewarding actions. For training more effective agents, we propose a framework that supports learning a flexible and well-regularized policy, which consists of a fully implicit policy and a regularization through the state-action visitation frequency induced by the current policy and that induced by the data-collecting behavior policy. We theoretically show the equivalence between policy-matching and state-action-visitation matching, and thus the compatibility of many prior work with our framework. An effective instantiation of our framework through the GAN structure is provided, together with some techniques to explicitly smooth the state-action mapping for robust generalization beyond the static dataset. Extensive experiments and ablation study on the D4RL dataset validate our framework and the effectiveness of our algorithmic designs.

📄 PDF Abstract BibTeX

Code (0)

등록된 구현이 없습니다.

Tasks

D4RLreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Similar Papers 제목 키워드 기반

A Behavior Regularized Implicit Policy for Offline Reinforcement Learning

2022-02-19 · Shentao Yang, Zhendong Wang, Huangjie Zheng, Yihao Feng 외

Offline reinforcement learning enables learning from a fixed dataset, without further interactions with the environment. The lack of environmental interactions makes the policy training vulnerable to state-action pairs f…

D4RLreinforcement-learningReinforcement LearningReinforcement Learning (RL)

Implicitly Regularized RL with Implicit Q-Values

2021-08-16 · Nino Vieillard, Marcin Andrychowicz, Anton Raichuk, Olivier Pietquin 외

The $Q$-function is a central quantity in many Reinforcement Learning (RL) algorithms for which RL agents behave following a (soft)-greedy policy w.r.t. to $Q$. It is a powerful tool that allows action selection without …

Reinforcement Learning (RL)

Implicit Policy for Reinforcement Learning

2018-06-10 · Yunhao Tang, Shipra Agrawal

We introduce Implicit Policy, a general class of expressive policies that can flexibly represent complex action distributions in reinforcement learning, with efficient algorithms to compute entropy regularized policy gra…

reinforcement-learningReinforcement LearningReinforcement Learning (RL)

Flow-Based Policy for Online Reinforcement Learning

2025-06-15 · Lei Lv, Yunfei Li, Yu Luo, Fuchun Sun 외

We present \textbf{FlowRL}, a novel framework for online reinforcement learning that integrates flow-based policy representation with Wasserstein-2-regularized optimization. We argue that in addition to training signals,…

reinforcement-learningReinforcement Learning

A unified view of entropy-regularized Markov decision processes

2017-05-22 · Gergely Neu, Anders Jonsson, Vicenç Gómez

We propose a general framework for entropy-regularized average-reward reinforcement learning in Markov decision processes (MDPs). Our approach is based on extending the linear-programming formulation of policy optimizati…

Policy Gradient Methodsreinforcement-learningReinforcement LearningReinforcement Learning (RL)