Inverse Preference Learning: Preference-based RL without a Reward Function.
Joey HejnaDorsa SadighPublished in: CoRR (2023)
Keyphrases
- reward function
- preference learning
- reinforcement learning
- reinforcement learning algorithms
- markov decision processes
- gaussian processes
- optimal policy
- state space
- ordinal regression
- pairwise comparison
- inverse reinforcement learning
- recommender systems
- multiple agents
- active learning
- preference relations
- user preferences
- dynamic programming
- ranking functions
- learning algorithm
- transition probabilities
- model free
- multi agent
- gaussian process
- multi objective
- training data
- temporal difference
- state variables
- objective function