Self-adaptive lower confidence bound: A new general and effective prescreening method for Gaussian Process surrogate model assisted evolutionary algorithms.
Bo LiuQingfu ZhangFrancisco V. FernándezGeorges G. E. GielenPublished in: IEEE Congress on Evolutionary Computation (2012)
Keyphrases
- gaussian process
- probabilistic model
- prior knowledge
- similarity measure
- objective function
- dynamical model
- evolutionary algorithm
- fully bayesian
- bayesian framework
- prior information
- em algorithm
- parameter estimation
- autoregressive
- energy function
- input data
- gaussian process regression
- sparse approximation
- hyperparameters
- pairwise
- learning algorithm
- closed form
- higher order
- latent variables
- probability distribution
- support vector
- feature selection