An extreme learning machine based virtual sample generation method with feature engineering for credit risk assessment with data scarcity
作者：
Highlights：
• An ELM based virtual sample generation method with feature engineering is proposed.
• ELM is utilized to generate virtual samples for solving data instance scarcity issue.
• Feature construction and feature selection are proposed to solve data attribute scarcity issue.
• The proposed method is used for credit classification with data scarcity.
• The proposed method outperforms the benchmark models in most cases.
摘要
•An ELM based virtual sample generation method with feature engineering is proposed.•ELM is utilized to generate virtual samples for solving data instance scarcity issue.•Feature construction and feature selection are proposed to solve data attribute scarcity issue.•The proposed method is used for credit classification with data scarcity.•The proposed method outperforms the benchmark models in most cases.
论文关键词：Extreme learning machine,Virtual sample generation,Data scarcity,Credit risk assessment
论文评审过程：Received 20 July 2021, Revised 6 January 2022, Accepted 25 April 2022, Available online 2 May 2022, Version of Record 4 May 2022.
论文官网地址：https://doi.org/10.1016/j.eswa.2022.117363
