Two-stage binary classifier with fuzzy-valued loss function.
Robert BurdukMarek KurzynskiPublished in: Pattern Anal. Appl. (2006)
Keyphrases
- loss function
- support vector
- empirical risk
- bayes rule
- binary classifiers
- boosting framework
- pairwise
- fuzzy sets
- learning to rank
- training data
- support vector machine
- error correcting output codes
- risk minimization
- logistic regression
- svm classifier
- reproducing kernel hilbert space
- feature space
- binary classification
- regularization term
- multi class
- t norm
- training samples
- feature selection
- solution path
- feature set
- classification algorithm
- multiple kernel learning
- ensemble learning
- cross validation
- kernel function