Model-Based Offline Reinforcement Learning with Pessimism-Modulated Dynamics Belief.
Kaiyang GuoYunfeng ShaoYanhui GengPublished in: CoRR (2022)
Keyphrases
- reinforcement learning
- model free
- function approximation
- state space
- dynamical systems
- markov decision processes
- dynamic model
- reinforcement learning algorithms
- belief revision
- multi agent
- policy search
- reinforcement learning methods
- temporal difference
- learning algorithm
- learning process
- robot control
- data driven
- transition model
- belief functions
- belief state
- transfer learning
- partially observable
- belief space
- supervised learning
- subjective logic