Asymmetric Beta Loss for Evidence-Based Safe Semi-Supervised Multi-Label Learning.
Hao-Zhe LiuMing-Kun XieChen-Chen ZongSheng-Jun HuangPublished in: KDD (2024)
Keyphrases
- multi label learning
- semi supervised
- labeled data
- label propagation
- semi supervised learning
- unlabeled data
- multi label
- pairwise
- supervised learning
- active learning
- labeled examples
- pairwise constraints
- transfer learning
- scene classification
- weakly supervised
- feature extraction
- text classification
- training data
- machine learning
- negative matrix factorization
- feature selection