Preventing Over-Fitting of Cross-Validation with Kernel Stability.
Yong LiuShizhong LiaoPublished in: ECML/PKDD (2) (2014)
Keyphrases
- cross validation
- support vector
- model selection
- rbf kernel
- kernel function
- hyperparameters
- variable selection
- regression problems
- kernel methods
- cross validated
- generalization error
- kernel machines
- unseen data
- information criterion
- training set
- kernel matrix
- parameter estimation
- leave one out cross validation
- error estimates
- regularization parameter
- feature space
- loss function
- em algorithm
- least squares
- semi supervised
- support vector machine
- data sets
- nearest neighbor classifiers
- database