Semi-supervised multi-class Adaboost by exploiting unlabeled data.
Enmin SongDongshan HuangGuangzhi MaChih-Cheng HungPublished in: Expert Syst. Appl. (2011)
Keyphrases
- multi class
- unlabeled data
- semi supervised
- pairwise
- labeled data
- semi supervised learning
- semi supervised classification
- labeled and unlabeled data
- co training
- multiclass classification
- multi class classification
- labeled examples
- object detection
- active learning
- support vector machine
- supervised learning
- base classifiers
- label propagation
- feature selection
- unsupervised learning
- pairwise constraints
- binary classification
- metric learning
- training examples
- multiple instance learning
- learning algorithm
- binary classifiers
- boosting algorithms
- text categorization
- ensemble learning
- cost sensitive
- improve the classification accuracy
- training data
- error correcting output codes
- class distribution
- transfer learning
- text classification
- unlabeled samples
- transductive learning
- training set
- binary classification problems
- class labels
- loss function
- high dimensional
- data sets
- data points
- learning process
- data mining