Boosting for learning from multiclass data sets via a regularized loss function.
Mohamed AbouelenienXiaohui YuanPublished in: GrC (2013)
Keyphrases
- multi class
- loss function
- boosting algorithms
- boosting framework
- data sets
- hinge loss
- learning algorithm
- pairwise
- learning models
- multiclass classification
- base classifiers
- feature selection
- multi task
- support vector machine
- supervised learning
- multi class classification
- learning tasks
- multiclass learning
- multi class boosting
- error correcting output codes
- risk minimization
- binary classification
- cost sensitive
- learning problems
- kernel methods
- decision boundary
- multi task learning
- active learning
- machine learning algorithms
- training set