Universum Selection for Boosting the Performance of Multiclass Support Vector Machines Based on One-versus-One Strategy.
Patoomsiri SongsiriVladimir CherkasskyBoonserm KijsirikulPublished in: Knowl. Based Syst. (2018)
Keyphrases
- multiclass support vector machines
- multi class
- binary classification
- cost sensitive
- misclassification costs
- semi supervised learning
- boosting algorithms
- multi class classification
- base classifiers
- multiclass problems
- unbalanced data
- maximum margin
- binary classification problems
- feature selection
- multi task
- support vector machine
- machine learning
- semi supervised
- semi definite programming
- binary classifiers
- training data
- learning algorithm
- unlabeled data
- ensemble methods
- unsupervised learning
- learning problems