When can unlabeled data improve the learning rate?
Christina GöpfertShai Ben-DavidOlivier BousquetSylvain GellyIlya O. TolstikhinRuth UrnerPublished in: CoRR (2019)
Keyphrases
- unlabeled data
- learning rate
- semi supervised learning
- labeled data
- learning algorithm
- semi supervised
- active learning
- co training
- improve the classification accuracy
- semi supervised classification
- text classification
- training data
- unlabeled training data
- labeled training data
- supervised learning
- labeled examples
- convergence rate
- number of labeled examples
- domain adaptation
- labeled and unlabeled data
- data points
- transfer learning
- training set
- text categorization
- training examples
- labeled data for training
- class labels
- machine learning
- learning tasks
- prior knowledge
- object recognition
- real world