BiES: Adaptive Policy Optimization for Model-Based Offline Reinforcement Learning.
Yijun YangJing JiangZhuowei WangQiqi DuanYuhui ShiPublished in: AI (2022)
Keyphrases
- reinforcement learning
- optimal policy
- model free
- policy search
- markov decision process
- optimization algorithm
- actor critic
- optimization problems
- markov decision processes
- reinforcement learning algorithms
- data driven
- optimal control
- state space
- function approximators
- control policy
- function approximation
- global optimization
- approximate dynamic programming
- policy gradient
- control policies
- policy iteration
- adaptive control
- policy evaluation
- action space
- machine learning
- reinforcement learning problems
- state and action spaces
- reward function
- action selection
- constrained optimization
- optimization method
- multi agent
- real time
- partially observable markov decision processes
- infinite horizon
- average reward
- decision problems
- dynamic programming
- search space
- genetic algorithm
- partially observable domains