Work

State Advantage Weighting for Offline RL

Year: 2022

Type: preprint

Abstract: We present state advantage weighting for offline reinforcement learning (RL). In contrast to action advantage $A(s,a)$ that we commonly adopt in QSA learning, we leverage state advantage $A(s,s^\prime)$ and QSS learning for offline RL, hence decoupling the action from values. We expect the agent can get to the high-reward state and the action is determined by how the agent can get to that corresponding state. Experiments on D4RL datasets show that our proposed method can achieve remarkable performance against the common baselines. Furthermore, our method shows good generalization capability when transferring from offline to online.

Source: arXiv (Cornell University)

DOI: 10.48550/arxiv.2210.04251

Authors: Lyu, Jiafei, Gong, Aicheng, Wan, Le, Lu, Zongqing, Li, Xiu

Institutions: Tsinghua Shenzhen International Graduate School, Tsinghua University, Peking University, Tencent (China)

Language: English

Topic: Reinforcement Learning in Robotics

Subfield: Artificial Intelligence

Field: Computer Science

Domain: Physical Sciences

Open Access status: green

OpenAlex ID: W4304701438

Cites:

Cited by: 2

Related to: 10