Regret Bounds for Gaussian Process Bandit Problems.
Steffen GrünewälderJean-Yves AudibertManfred OpperJohn Shawe-TaylorPublished in: AISTATS (2010)
Keyphrases
- gaussian process
- bandit problems
- regret bounds
- decision problems
- lower bound
- linear regression
- bayesian framework
- online learning
- regression model
- approximate inference
- model selection
- multi armed bandit
- semi supervised
- hyperparameters
- latent variables
- upper bound
- closed form
- dynamic programming
- maximum a posteriori
- random sampling
- probabilistic model
- bregman divergences
- special case
- objective function