M2SPL: Generative multiview features with adaptive meta-self-paced sampling for class-imbalance learning

作者：

Highlights：

• A multi-view adaptive meta self-paced learning is presented for imbalanced classification.

• A multiple feature selection approach is used to generate multiview features.

• M2SPL performs multiple high-quality sampling procedures in multiple views.

• The performance of M2SPL is higher than other competitors, which demonstrates the superiority of our method.

摘要

•A multi-view adaptive meta self-paced learning is presented for imbalanced classification.•A multiple feature selection approach is used to generate multiview features.•M2SPL performs multiple high-quality sampling procedures in multiple views.•The performance of M2SPL is higher than other competitors, which demonstrates the superiority of our method.

论文关键词：Imbalance learning,Multiview generation,Adaptive sampling,Meta learner,Self-paced learning

论文评审过程：Received 1 March 2021, Revised 17 May 2021, Accepted 28 September 2021, Available online 14 October 2021, Version of Record 29 October 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115999