DeepAI
Log In Sign Up

State Advantage Weighting for Offline RL

10/09/2022
by   Jiafei Lyu, et al.
0

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^') 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.

READ FULL TEXT

page 1

page 2

page 3

page 4

06/07/2021

Believe What You See: Implicit Constraint Approach for Offline Multi-Agent Reinforcement Learning

Learning from datasets without interaction with environments (Offline Le...
07/05/2021

The Least Restriction for Offline Reinforcement Learning

Many practical applications of reinforcement learning (RL) constrain the...
05/27/2019

LAW: Learning to Auto Weight

Example weighting algorithm is an effective solution to the training bia...
10/12/2022

Semi-Supervised Offline Reinforcement Learning with Action-Free Trajectories

Natural agents can effectively learn from multiple data sources that dif...
05/29/2019

Advantage Amplification in Slowly Evolving Latent-State Environments

Latent-state environments with long horizons, such as those faced by rec...
09/30/2022

S2P: State-conditioned Image Synthesis for Data Augmentation in Offline Reinforcement Learning

Offline reinforcement learning (Offline RL) suffers from the innate dist...
07/01/2021

Offline-to-Online Reinforcement Learning via Balanced Replay and Pessimistic Q-Ensemble

Recent advance in deep offline reinforcement learning (RL) has made it p...