Model-Based Offline Reinforcement Learning with Local Misspecification.
Kefan DongYannis Flet-BerliacAllen NieEmma BrunskillPublished in: AAAI (2023)
Keyphrases
- reinforcement learning
- model free
- function approximation
- real time
- reinforcement learning algorithms
- machine learning
- function approximators
- state space
- optimal policy
- markov decision processes
- control problems
- temporal difference
- temporal difference learning
- database
- action selection
- optimal control
- data driven
- least squares
- dynamic programming
- learning process
- multi agent
- bayesian networks
- case study
- knowledge base