Minority manifold regularization by stacked auto-encoder for imbalanced learning

作者：

Highlights：

• A novel imbalanced learning method based on stacked regularized auto-encoders.

• Regularizers are adapted from manifold learning and fisher discriminant analysis.

• Captures underlying manifold of minority class through one-class learning approach.

• Results over 20 datasets confirm the effectiveness of the proposed method.

摘要

•A novel imbalanced learning method based on stacked regularized auto-encoders.•Regularizers are adapted from manifold learning and fisher discriminant analysis.•Captures underlying manifold of minority class through one-class learning approach.•Results over 20 datasets confirm the effectiveness of the proposed method.

论文关键词：Regularized auto-encoder,Imbalanced data classification,Feature learning

论文评审过程：Received 23 November 2019, Revised 11 November 2020, Accepted 12 November 2020, Available online 17 November 2020, Version of Record 10 February 2021.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114317