Guarantees for Epsilon-Greedy Reinforcement Learning with Function Approximation.
Christoph DannYishay MansourMehryar MohriAyush SekhariKarthik SridharanPublished in: ICML (2022)
Keyphrases
- function approximation
- reinforcement learning
- temporal difference learning
- temporal difference
- tile coding
- function approximators
- model free
- temporal difference learning algorithms
- mountain car
- radial basis function
- dynamic programming
- reinforcement learning algorithms
- state action space
- learning tasks
- state space
- td learning
- feature selection
- markov decision processes
- control problems
- learning problems
- transfer learning
- action selection
- search space
- training set
- support vector
- machine learning