On the Sample Complexity and Metastability of Heavy-tailed Policy Search in Continuous Control.
Amrit Singh BediAnjaly ParayilJunyu ZhangMengdi WangAlec KoppelPublished in: CoRR (2021)
Keyphrases
- sample complexity
- policy search
- heavy tailed
- learning problems
- theoretical analysis
- learning algorithm
- special case
- active learning
- upper bound
- pac learning
- supervised learning
- lower bound
- generalization error
- training examples
- generalized gaussian
- continuous state
- machine learning
- sample size
- reinforcement learning
- state space
- objective function