A review of methods for imbalanced multi-label classification

作者：

Highlights：

• Three types of imbalanced problems are common challenges in multi-label classification: imbalance within labels, between labels, and among label-sets.

• A comprehensive and up-to-date review of methods for addressing imbalanced problems in multi-label classification is presented.

• Methods for assessing imbalance level and performance measures in the multi-label scenario are surveyed.

• Comparative analysis of the reviewed methods and their limitations are discussed to guide future directions.

摘要

•Three types of imbalanced problems are common challenges in multi-label classification: imbalance within labels, between labels, and among label-sets.•A comprehensive and up-to-date review of methods for addressing imbalanced problems in multi-label classification is presented.•Methods for assessing imbalance level and performance measures in the multi-label scenario are surveyed.•Comparative analysis of the reviewed methods and their limitations are discussed to guide future directions.

论文关键词：Imbalanced Data,Multi-label Classification,Imbalanced Classification,Machine learning,Imbalanced Approaches,Review on Imbalanced Classification

论文评审过程：Received 29 January 2020, Revised 18 March 2021, Accepted 26 March 2021, Available online 6 May 2021, Version of Record 16 May 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.107965