In-sample Model Selection for Trimmed Hinge Loss Support Vector Machine.
Davide AnguitaAlessandro GhioLuca OnetoSandro RidellaPublished in: Neural Process. Lett. (2012)
Keyphrases
- model selection
- soft margin
- support vector machine
- sample size
- parameter selection
- generalization error
- cross validation
- decision boundary
- feature selection
- machine learning
- parameter estimation
- hyperparameters
- support vector
- regression model
- multi class
- svm classifier
- loss function
- generalization ability
- gaussian process
- boosting algorithms
- variable selection
- binary classification
- feature space
- linear classifiers
- mixture model
- random sampling
- likelihood function
- kernel function
- training set
- data mining
- support vector machine svm