The LoCA Regret: A Consistent Metric to Evaluate Model-Based Behavior in Reinforcement Learning.
Harm van SeijenHadi NekoeiEvan RacahSarath ChandarPublished in: CoRR (2020)
Keyphrases
- reinforcement learning
- model free
- function approximation
- state space
- supervised learning
- metric space
- distance measure
- real robot
- reinforcement learning algorithms
- lower bound
- loss function
- distance function
- distance metric
- worst case
- mobile robot
- metric learning
- autonomous robots
- action selection
- multi agent
- knowledge base