!MDP Playground: Meta-Features in Reinforcement Learning.
Raghu RajanFrank HutterPublished in: CoRR (2019)
Keyphrases
- reinforcement learning
- markov decision processes
- markov decision process
- optimal policy
- state space
- learning algorithm
- feature set
- feature vectors
- function approximation
- image features
- feature space
- linear programming
- object recognition
- feature extraction
- low level
- false positives
- meta level
- reward function
- initial state
- machine learning
- factored markov decision processes