ALLSH: Active Learning Guided by Local Sensitivity and Hardness.
Shujian ZhangChengyue GongXingchao LiuPengcheng HeWeizhu ChenMingyuan ZhouPublished in: NAACL-HLT (Findings) (2022)
Keyphrases
- active learning
- random sampling
- machine learning
- selective sampling
- sensitivity analysis
- learning algorithm
- computational complexity
- learning process
- batch mode
- semi supervised
- high sensitivity
- learning strategies
- imbalanced data classification
- learning theory
- np complete
- neural network
- music emotion classification
- relevance feedback
- worst case
- transfer learning
- unlabeled data
- labeled data
- phase transition
- class imbalance
- sample complexity
- np hardness
- semi supervised learning
- pairwise
- sample selection
- pool based active learning
- information systems