Reinforcement learning from comparisons: Three alternatives is enough, two is not
Benoit LaslierJean-François LaslierPublished in: CoRR (2013)
Keyphrases
- reinforcement learning
- function approximation
- temporal difference
- machine learning
- state space
- reinforcement learning algorithms
- action selection
- alternative approaches
- markov decision processes
- model free
- reinforcement learning methods
- multiple criteria
- optimal control
- evaluation function
- optimal policy
- action space
- function approximators
- computational complexity