FARE: Enabling Fine-grained Attack Categorization under Low-quality Labeled Data.
Junjie LiangWenbo GuoTongbo LuoVasant HonavarGang WangXinyu XingPublished in: NDSS (2021)
Keyphrases
- fine grained
- low quality
- labeled data
- unlabeled data
- semi supervised
- semi supervised learning
- text categorization
- high quality
- coarse grained
- text classification
- active learning
- training data
- supervised learning
- access control
- transfer learning
- class labels
- domain adaptation
- labeled training data
- data points
- co training
- labeled and unlabeled data
- learning algorithm
- training examples
- prior knowledge
- pairwise
- unsupervised learning
- face recognition
- machine learning
- labeling effort
- data lineage
- labeled data for training
- palmprint
- target domain
- feature selection