Regularization Guarantees Generalization in Bayesian Reinforcement Learning through Algorithmic Stability.
Aviv TamarDaniel SoudryEv ZisselmanPublished in: AAAI (2022)
Keyphrases
- generalization bounds
- algorithmic stability
- bayesian reinforcement learning
- data dependent
- learning theory
- generalization error
- uniform convergence
- optimal policy
- reinforcement learning
- generalization ability
- monte carlo tree search
- model selection
- machine learning
- learning problems
- kernel machines
- risk minimization
- statistical learning theory
- vc dimension
- learning algorithm
- ranking functions
- reproducing kernel hilbert space
- state space
- learning machines
- upper bound
- ranking algorithm
- markov decision processes
- linear classifiers
- supervised learning
- dynamic programming
- semi supervised learning
- active learning
- lower bound
- data mining