Compact learning for multi-label classification

作者：

Highlights：

• We propose a new framework named CL to embed the features and labels simultaneously.

• The embedding processes of the two space in CL are with mutual guidance.

• We propose a simple yet effective instantiation of CL named CMLL.

• CMLL is compatible with flexible multi-label classifiers.

摘要

•We propose a new framework named CL to embed the features and labels simultaneously.•The embedding processes of the two space in CL are with mutual guidance.•We propose a simple yet effective instantiation of CL named CMLL.•CMLL is compatible with flexible multi-label classifiers.

论文关键词：Machine learning,Multi-label classification,Label compression,Compact learning

论文评审过程：Received 29 November 2019, Revised 22 November 2020, Accepted 16 January 2021, Available online 21 January 2021, Version of Record 28 January 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.107833