Automated imbalanced classification via meta-learning

作者：

Highlights：

• The first automated machine learning approach to imbalanced domains.

• 95% reduction of computational efforts concerning workflow selection and optimisation.

• Competitive results in comparison to state-of-the-art AutoML methods.

摘要

•The first automated machine learning approach to imbalanced domains.•95% reduction of computational efforts concerning workflow selection and optimisation.•Competitive results in comparison to state-of-the-art AutoML methods.

论文关键词：Imbalance domain learning,Automated machine learning,Meta-learning,Classification

论文评审过程：Received 31 December 2020, Revised 27 March 2021, Accepted 3 April 2021, Available online 20 April 2021, Version of Record 24 April 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.115011