Adaptable Conservative Q-Learning for Offline Reinforcement Learning.
Lyn QiuXu LiLenghan LiangMingming SunJunchi YanPublished in: PRCV (3) (2023)
Keyphrases
- reinforcement learning
- function approximation
- reinforcement learning algorithms
- model free
- state space
- learning algorithm
- temporal difference learning
- action selection
- continuous state and action spaces
- state action space
- optimal policy
- multi agent
- multi agent reinforcement learning
- stochastic approximation
- markov decision processes
- real time
- temporal difference
- reward function
- optimal control
- control problems
- learning agent
- continuous state
- rl algorithms
- learning process
- eligibility traces
- policy search
- supervised learning
- markov decision problems
- reinforcement learning methods
- partially observable
- cooperative
- policy evaluation
- learning agents
- policy iteration
- relational reinforcement learning
- real robot
- fixed point