On the Minimal Supervision for Training Any Binary Classifier from Only Unlabeled Data.
Nan LuGang NiuAditya Krishna MenonMasashi SugiyamaPublished in: ICLR (Poster) (2019)
Keyphrases
- unlabeled data
- labeled training data
- labeled data
- training set
- active learning
- training examples
- semi supervised learning
- training data
- supervised learning
- co training
- number of labeled examples
- semi supervised
- labeled instances
- classifier training
- labeled and unlabeled data
- text classification
- labeled data for training
- training and test data
- unlabeled examples
- class labels
- text classifiers
- labeled examples
- training and test sets
- training process
- sample selection bias
- semi supervised classification
- training samples
- text categorization
- learning algorithm
- decision boundary
- semi supervised learning methods
- partially labeled data
- labeling effort
- prior knowledge
- unsupervised learning
- active learner
- class distribution
- decision trees
- transfer learning
- data points
- machine learning
- unlabeled samples
- feature selection
- positive examples
- supervised methods
- feature space
- unlabeled instances
- multi view
- naive bayes
- support vector
- classification accuracy
- binary classifiers
- svm classifier
- data sets
- domain adaptation
- training instances