Model Selection Using a Class of Kernels with an Invariant Metric.
Akira TanakaMasashi SugiyamaHideyuki ImaiMineichi KudoMasaaki MiyakoshiPublished in: SSPR/SPR (2006)
Keyphrases
- model selection
- cross validation
- learning machines
- hyperparameters
- bayesian learning
- sample size
- variable selection
- statistical inference
- mixture model
- regression model
- model selection criteria
- gaussian process
- feature selection
- generalization bounds
- selection criterion
- machine learning
- parameter estimation
- statistical learning
- bayesian methods
- gaussian processes
- generalization error
- bayesian model selection
- feature space
- marginal likelihood
- error estimation
- leave one out cross validation
- automatic model selection
- information criterion
- support vector
- kernel machines
- support vector machine
- kernel function