Stable Control Policy and Transferable Reward Function via Inverse Reinforcement Learning.
Keyu WuFengge WuYijun LinJunsuo ZhaoPublished in: ICCAI (2023)
Keyphrases
- inverse reinforcement learning
- control policy
- reward function
- reinforcement learning
- control policies
- approximate dynamic programming
- long run
- markov decision processes
- optimal policy
- preference elicitation
- reinforcement learning algorithms
- state space
- average cost
- partially observable
- function approximation
- learning algorithm
- multiple agents
- transition probabilities
- temporal difference
- model free
- multi agent
- dynamic programming
- control system
- finite horizon
- action space
- finite state
- utility function