Creating synthetic minority class samples based on autoencoder extreme learning machine

作者：

Highlights：

• A novel synthetic sample creation mechanism for imbalanced learning is proposed.

• We create the synthetic samples with autoencoder extreme learning machine.

• The crossover, mutation and filtration operations are used to create the synthetic samples.

• We conduct the exhaustive experiments to demonstrate effectiveness of our method.

摘要

•A novel synthetic sample creation mechanism for imbalanced learning is proposed.•We create the synthetic samples with autoencoder extreme learning machine.•The crossover, mutation and filtration operations are used to create the synthetic samples.•We conduct the exhaustive experiments to demonstrate effectiveness of our method.

论文关键词：Imbalanced classification,Minority class,Majority class,Synthetic samples creation,SMOTE

论文评审过程：Received 14 December 2020, Revised 9 June 2021, Accepted 4 July 2021, Available online 20 July 2021, Version of Record 31 July 2021.

论文官网地址：https://doi.org/10.1016/j.patcog.2021.108191