Penalty and margin decomposition - an inspection of loss function regularization in SVM.
Wei-Chih LinChan-Yun YangGene Eu JanJr-Syu YangPublished in: ICNSC (2015)
Keyphrases
- loss function
- hinge loss
- support vector
- risk minimization
- regularization term
- reproducing kernel hilbert space
- stochastic gradient descent
- convex loss functions
- square loss
- pairwise
- logistic regression
- learning to rank
- norm regularization
- solution path
- gradient boosting
- empirical risk
- boosting algorithms
- objective function
- maximum margin
- binary classification
- multi class
- kernel function
- soft margin
- svm classifier
- decision boundary
- hyperplane
- kernel methods
- support vector machine svm
- kernel machines
- pairwise constraints
- potential functions
- support vectors
- learning problems