Regret Bound Balancing and Elimination for Model Selection in Bandits and RL.
Aldo PacchianoChristoph DannClaudio GentilePeter L. BartlettPublished in: CoRR (2020)
Keyphrases
- regret bounds
- model selection
- reinforcement learning
- online learning
- lower bound
- cross validation
- linear regression
- parameter estimation
- hyperparameters
- upper bound
- machine learning
- sample size
- mixture model
- feature selection
- model selection criteria
- gaussian process
- bregman divergences
- generalization error
- data mining
- mutual information
- closed form
- state space
- objective function