Don't fear the unlabelled: safe semi-supervised learning via debiasing.
Hugo SchmutzOlivier HumbertPierre-Alexandre MatteiPublished in: ICLR (2023)
Keyphrases
- semi supervised learning
- unlabeled data
- semi supervised
- labeled data
- supervised learning
- semi supervised classification
- machine learning
- training data
- unsupervised learning
- label propagation
- text categorization
- learning problems
- learning models
- graph based semi supervised learning
- semi supervised learning methods
- co training
- transfer learning
- active learning
- labeled examples
- graph construction
- regularization framework
- manifold regularization
- domain adaptation
- semi supervised learning algorithms
- data sets
- prior knowledge
- feature selection