A Sample Selection Approach for Universal Domain Adaptation.
Omri LifshitzLior WolfPublished in: CoRR (2020)
Keyphrases
- domain adaptation
- sample selection
- active learning
- training data
- labeled data
- semi supervised
- unlabeled data
- semi supervised learning
- multiple sources
- selection strategy
- target domain
- cross domain
- transfer learning
- test data
- support vector machine
- data sets
- document classification
- training samples
- machine learning
- test set
- training examples
- learning algorithm
- sentiment classification
- supervised learning
- classification accuracy
- databases
- co training
- error rate
- neural network
- data model
- training set
- k nearest neighbor
- learning process
- data sources