Sample Complexity and Overparameterization Bounds for Projection-Free Neural TD Learning.
Semih CayciSiddhartha SatpathiNiao HeR. SrikantPublished in: CoRR (2021)
Keyphrases
- sample complexity
- td learning
- vc dimension
- upper bound
- lower bound
- temporal difference
- evaluation function
- covering numbers
- theoretical analysis
- learning problems
- supervised learning
- function approximation
- pac learning
- average case
- learning algorithm
- generalization error
- active learning
- reinforcement learning
- special case
- worst case
- reinforcement learning algorithms
- neural network
- sample size
- optimal solution
- training data
- machine learning
- learning tasks
- model free
- monte carlo
- model selection
- learning process