An extensive study of C-SMOTE, a Continuous Synthetic Minority Oversampling Technique for Evolving Data Streams

作者：

Highlights：

• There is a trade-off between the performances of the two classes.

• The performances of the minority class are, in most of the cases, increased.

• The gain in the minority class recall is bigger than the loss in the majority one.

• Time and RAM consumed are more than the state-of-the-art ones.

摘要

•There is a trade-off between the performances of the two classes.•The performances of the minority class are, in most of the cases, increased.•The gain in the minority class recall is bigger than the loss in the majority one.•Time and RAM consumed are more than the state-of-the-art ones.

论文关键词：Evolving Data Stream,Streaming,Concept drift,Balancing

论文评审过程：Received 14 December 2020, Revised 25 January 2022, Accepted 30 January 2022, Available online 12 February 2022, Version of Record 17 February 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.116630