A safe sample screening rule for Universum support vector machines.
Jiang ZhaoYitian XuPublished in: Knowl. Based Syst. (2017)
Keyphrases
- binary classification
- support vector
- maximum margin
- learning machines
- semi supervised learning
- labeled samples
- multi class
- hyperplane
- multi class classification
- support vector machine
- semi supervised
- semi definite programming
- cost sensitive
- multi label
- unbalanced data
- kernel methods
- kernel function
- cross validation
- markov networks
- training examples
- feature selection
- generalization error
- learning algorithm
- support vector regression
- learning problems
- loss function
- pattern classification
- model selection
- data sets
- ensemble methods
- support vectors
- sample size
- multiple kernel learning
- misclassification costs
- decision function
- prediction accuracy