An ensemble contrastive classification framework for imbalanced learning with sample-neighbors pair construction
作者：
Highlights：
• The imbalanced classification is redefined as a label matching task.
• Sample-neighbors pairs is constructed to balance classes without introducing noise.
• An ensemble comparative classification framework is presented for robust effect.
摘要
•The imbalanced classification is redefined as a label matching task.•Sample-neighbors pairs is constructed to balance classes without introducing noise.•An ensemble comparative classification framework is presented for robust effect.
论文关键词：Imbalanced classification,Contrastive learning,Sample-neighbors pair construction,Ensemble classification framework
论文评审过程：Received 20 January 2022, Revised 4 May 2022, Accepted 5 May 2022, Available online 13 May 2022, Version of Record 18 May 2022.
论文官网地址：https://doi.org/10.1016/j.knosys.2022.109007
