An imbalanced binary classification method based on contrastive learning using multi-label confidence comparisons within sample-neighbors pair.
Xin GaoZhihang MengXin JiaJing LiuXinping DiaoBing XueZijian HuangKangsheng LiPublished in: Neurocomputing (2023)
Keyphrases
- classification method
- multi label
- text classification
- learning tasks
- learning algorithm
- learning process
- multi label learning
- prior knowledge
- graph cuts
- associative classification
- support vector machine svm
- text categorization
- active learning
- classification algorithm
- small number
- markov random field
- nearest neighbor
- support vector machine
- multi label classification
- feature selection