Benchmarking state-of-the-art imbalanced data learning approaches for credit scoring
作者：
Highlights：
• Proposal of a new taxonomy for imbalanced data learning approaches (IDLAs).
• Large-scale benchmark of 28 IDLAs across three real-world credit scoring datasets.
• Assessment of the performance of generative adversarial nets in credit scoring.
• Analysis of the internal cause of class imbalance problem in credit scoring.
• Recommendation for the selection strategy of IDLAs in credit scoring.
摘要
•Proposal of a new taxonomy for imbalanced data learning approaches (IDLAs).•Large-scale benchmark of 28 IDLAs across three real-world credit scoring datasets.•Assessment of the performance of generative adversarial nets in credit scoring.•Analysis of the internal cause of class imbalance problem in credit scoring.•Recommendation for the selection strategy of IDLAs in credit scoring.
论文关键词：Credit scoring,Imbalanced classification,Predicting benchmark,GAN
论文评审过程：Received 30 March 2022, Revised 26 August 2022, Accepted 18 September 2022, Available online 12 October 2022, Version of Record 17 October 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.118878
