Model Selection in Contextual Stochastic Bandit Problems.
Aldo PacchianoMy PhanYasin Abbasi-YadkoriAnup RaoJulian ZimmertTor LattimoreCsaba SzepesváriPublished in: CoRR (2020)
Keyphrases
- model selection
- bandit problems
- multi armed bandits
- cross validation
- decision problems
- hyperparameters
- parameter estimation
- bayesian learning
- regression model
- sample size
- error estimation
- variable selection
- machine learning
- selection criterion
- gaussian process
- statistical inference
- marginal likelihood
- generalization error
- bayesian methods
- monte carlo
- mixture model
- model selection criteria
- parameter determination
- automatic model selection
- mutual information
- evolutionary algorithm
- optimal strategy
- decision makers
- bayesian information criterion
- information criterion
- feature selection