On the Minimal Supervision for Training Any Binary Classifier from Only Unlabeled Data.
Nan LuGang NiuAditya Krishna MenonMasashi SugiyamaPublished in: CoRR (2018)
Keyphrases
- unlabeled data
- labeled training data
- training set
- labeled data
- active learning
- training examples
- semi supervised learning
- co training
- supervised learning
- semi supervised
- training data
- labeled instances
- labeled and unlabeled data
- number of labeled examples
- classifier training
- labeled data for training
- class labels
- unlabeled examples
- text classification
- training and test data
- labeled examples
- semi supervised classification
- training process
- text categorization
- training samples
- training and test sets
- labeling effort
- learning algorithm
- data points
- supervised methods
- class distribution
- sample selection bias
- prior knowledge
- decision boundary
- test data
- unlabeled samples
- training instances
- semi supervised learning methods
- text classifiers
- partially labeled data
- transfer learning
- feature space
- unlabeled instances
- machine learning
- positive examples
- data sets
- single view
- binary classifiers