Modified criterion to select useful unlabeled data for improving semi-supervised support vector machines.
Thanh-Binh LeSang-Woon KimPublished in: Pattern Recognit. Lett. (2015)
Keyphrases
- semi supervised support vector machines
- unlabeled data
- semi supervised
- semi supervised learning
- labeled data
- semi supervised classification
- co training
- labeled and unlabeled data
- supervised learning
- active learning
- text categorization
- unsupervised learning
- labeled examples
- pairwise
- machine learning
- multi view
- labeled training data
- number of labeled examples
- training examples
- class labels
- label propagation
- transfer learning
- supervised and semi supervised
- decision trees
- semisupervised learning
- small set of labeled
- training set
- data points
- e learning
- transductive learning
- supervised learning algorithms
- similarity measure
- text classification