The Use of Unlabeled Data versus Labeled Data for Stopping Active Learning for Text Classification.
Garrett BeattyEthan KochisMichael BloodgoodPublished in: CoRR (2019)
Keyphrases
- labeled data
- text classification
- unlabeled data
- active learning
- semi supervised learning
- unlabeled examples
- transfer learning
- unlabeled instances
- semi supervised classification
- machine learning
- labeled examples
- co training
- text categorization
- labeled and unlabeled data
- labeled training data
- class labels
- text mining
- training examples
- domain adaptation
- number of labeled examples
- feature selection
- text classifiers
- supervised learning algorithms
- naive bayes
- abundant unlabeled data
- unlabeled documents
- multi label
- semi supervised
- selective sampling
- labeled instances
- labeling effort
- learning algorithm
- label propagation
- sentiment classification
- labeling process
- positive examples
- query by committee
- document classification
- unsupervised learning
- training set
- supervised learning
- learning tasks
- training data
- transductive svm
- semisupervised learning
- semi supervised learning methods
- training and test data
- unlabeled samples
- pairwise constraints