Steady-State Error Compensation for Reinforcement Learning with Quadratic Rewards.
Liyao WangZishun ZhengYuan LinPublished in: CoRR (2024)
Keyphrases
- steady state
- reinforcement learning
- error compensation
- markov chain
- state space
- markov decision processes
- queue length
- product form
- operating conditions
- learning algorithm
- explicit expressions
- model free
- queueing model
- optimal policy
- computational complexity
- machine learning
- queueing networks
- reward function
- service times
- state dependent
- arrival rate
- objective function
- phase shifting
- structured light
- fluid model
- steady states
- variance estimator
- genetic regulatory networks
- mobile robot
- heavy traffic
- complex systems