Multi-Class Active Learning by Uncertainty Sampling with Diversity Maximization.
Yi YangZhigang MaFeiping NieXiaojun ChangAlexander G. HauptmannPublished in: Int. J. Comput. Vis. (2015)
Keyphrases
- uncertainty sampling
- multi class
- cost sensitive
- active learning
- ensemble members
- misclassification costs
- binary classification
- experimental design
- class distribution
- support vector machine
- multi class classification
- probability estimates
- pairwise
- multiple classes
- class imbalance
- feature selection
- random sampling
- multi class classifier
- base classifiers
- labeled data
- class probabilities
- supervised learning
- training set
- unlabeled data
- binary classifiers
- text classification
- learning process
- lower bound