A Light-Weight Multi-Objective Asynchronous Hyper-Parameter Optimizer.
Gabriel MaherStephen P. BoydMykel John KochenderferCristian MatacheAlex UlitskySlava YukhymukLeonid KopmanPublished in: CoRR (2022)
Keyphrases
- lightweight
- multi objective
- hyperparameters
- optimization algorithm
- model selection
- cross validation
- closed form
- evolutionary algorithm
- bayesian framework
- bayesian inference
- random sampling
- multi objective optimization
- support vector
- prior information
- em algorithm
- maximum likelihood
- sample size
- objective function
- parameter settings
- noise level
- incremental learning
- maximum a posteriori
- incomplete data
- genetic algorithm
- communication infrastructure
- parameter space
- missing values
- image reconstruction
- input data
- pairwise
- feature selection
- data sets