Parallelizing exploration-exploitation tradeoffs in Gaussian process bandit optimization.
Thomas DesautelsAndreas KrauseJoel W. BurdickPublished in: J. Mach. Learn. Res. (2014)
Keyphrases
- gaussian process
- exploration exploitation
- bandit problems
- gaussian processes
- active learning
- gaussian process regression
- regression model
- semi supervised
- hyperparameters
- latent variables
- random sampling
- bayesian framework
- model selection
- decision problems
- gaussian process models
- reinforcement learning
- data sets
- bayesian inference
- prior information
- information retrieval