Hyperparameter Search Space Pruning - A New Component for Sequential Model-Based Hyperparameter Optimization.
Martin WistubaNicolas SchillingLars Schmidt-ThiemePublished in: ECML/PKDD (2) (2015)
Keyphrases
- gaussian processes
- hyperparameters
- model selection
- gaussian process
- optimization algorithm
- grid search
- global optimization
- cross validation
- search space pruning
- bilevel programming
- closed form
- random sampling
- optimization process
- optimization problems
- maximum a posteriori
- maximum likelihood
- constrained optimization
- optimization model
- optimization strategies
- data driven
- artificial neural networks
- discrete optimization
- knowledge base
- data mining
- data sets