Towards the Transferability of Rewards Recovered via Regularized Inverse Reinforcement Learning.
Andreas SchlaginhaufenMaryam KamgarpourPublished in: CoRR (2024)
Keyphrases
- inverse reinforcement learning
- reward function
- bayesian nonparametric
- reinforcement learning
- partially observable environments
- markov decision processes
- state space
- optimal policy
- reinforcement learning algorithms
- multiple agents
- partially observable
- preference elicitation
- transition probabilities
- simple examples
- markov decision process
- temporal difference
- search space
- state variables
- hidden markov models