Geometric SMOTE for regression

作者：

Highlights：

• Imbalanced learning is a well-known problem in classification.

• Regression problems can also be affected by the imbalanced nature of data.

• We propose a method to mitigate the imbalanced regression problem.

• Improves the prediction of rare and extreme values of a continuous target variable.

摘要

•Imbalanced learning is a well-known problem in classification.•Regression problems can also be affected by the imbalanced nature of data.•We propose a method to mitigate the imbalanced regression problem.•Improves the prediction of rare and extreme values of a continuous target variable.

论文关键词：Imbalanced,Regression,Data-level

论文评审过程：Received 28 May 2021, Revised 21 September 2021, Accepted 7 December 2021, Available online 3 January 2022, Version of Record 18 January 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2021.116387