Boosting Source-free Domain Adaptation via Confidence-based Subsets Feature Alignment.
Hao-Wei YehThomas WestfechtelJia-Bin HuangTatsuya HaradaPublished in: ICPR (2022)
Keyphrases
- domain adaptation
- manifold alignment
- multiple sources
- target domain
- source domain
- cross domain
- transfer learning
- semi supervised
- labeled data
- feature representation
- covariate shift
- test data
- sentiment classification
- semi supervised learning
- feature vectors
- training data
- learning algorithm
- unlabeled data
- information retrieval
- support vector machine
- data model
- feature extraction
- data distribution
- training samples
- unsupervised learning
- document classification
- feature set
- image features
- information extraction
- co training
- data sources
- feature selection
- machine learning
- source and target domains