Combining binary-SVM and pairwise label constraints for multi-label classification.
Weifeng GuBenhui ChenJinglu HuPublished in: SMC (2010)
Keyphrases
- multi label classification
- binary classifiers
- multi class
- pairwise
- multi label
- binary classification
- max margin
- classification algorithm
- multi class classification
- semi supervised
- label space
- support vector
- support vector machine
- svm classifier
- pairwise constraints
- text categorization
- semi supervised learning
- hyperplane
- multiple labels
- support vector machine svm
- image classification
- knn
- image annotation
- graph cuts
- markov random field
- similarity measure
- higher order
- text classification
- cost sensitive
- neural network
- training data
- feature vectors
- kernel function
- cross validation
- image segmentation
- active learning
- feature selection
- multi label learning
- decision trees
- feature extraction
- ls svm
- prior knowledge
- generalization error
- learning problems
- learning tasks
- supervised learning
- prediction accuracy
- training examples