Not All Unlabeled Data are Equal: Learning to Weight Data in Semi-supervised Learning.
Zhongzheng RenRaymond A. YehAlexander G. SchwingPublished in: CoRR (2020)
Keyphrases
- unlabeled data
- semi supervised learning
- labeled data
- labeled and unlabeled data
- semi supervised
- supervised learning
- co training
- active learning
- training data
- learning models
- learning problems
- unsupervised learning
- semi supervised classification
- supervised learning algorithms
- partially labeled
- learning algorithm
- labeled training data
- labeled instances
- multi view learning
- data points
- text classification
- prior knowledge
- manifold regularization
- semi supervised learning setting
- domain adaptation
- positive examples
- training examples
- learning frameworks
- class labels
- number of labeled examples
- partially labeled data
- learning tasks
- training set
- data sets
- machine learning
- semi supervised learning methods
- unlabeled instances
- abundant unlabeled data
- improve the classification accuracy
- label propagation
- labeled examples
- transfer learning
- unlabeled examples
- learning process
- pairwise constraints
- weakly supervised
- multiple instance learning
- metric learning
- information extraction
- knowledge discovery
- reinforcement learning