Label shift conditioned hybrid querying for deep active learning.
Jiaqi LiHaojia KongGezheng XuChangjian ShuiRuizhi PuZhao KangCharles X. LingBoyu WangPublished in: Knowl. Based Syst. (2023)
Keyphrases
- active learning
- label noise
- active learning framework
- databases
- labeling effort
- unlabeled instances
- learning process
- hybrid learning
- multi label
- selective sampling
- machine learning
- learning strategies
- deep learning
- database
- query language
- supervised learning
- semi supervised
- hybrid approaches
- labeled data
- data model
- training data
- query by committee
- active learner
- image labeling
- experimental design
- random sampling
- transfer learning
- class labels
- unlabeled data
- training set
- feature space
- support vector
- feature selection
- data mining
- neural network