An enterprise default discriminant model based on optimal misjudgment loss ratio

作者：

Highlights：

• A 20-feature set has the highest discriminant accuracy.

• This set includes asset-liability ratio and total capitalization rate.

• The discriminant accuracy is highest when Type II loss–Type I loss ratio is 10.4.

• Our model offers the lowest Type II error rate and highest accuracy rate and G-mean.

摘要

•A 20-feature set has the highest discriminant accuracy.•This set includes asset-liability ratio and total capitalization rate.•The discriminant accuracy is highest when Type II loss–Type I loss ratio is 10.4.•Our model offers the lowest Type II error rate and highest accuracy rate and G-mean.

论文关键词：Default discriminant,Feature set selection,Misclassification cost,Lasso-logistic regression,Big data

论文评审过程：Received 30 November 2020, Revised 10 May 2022, Accepted 27 May 2022, Available online 31 May 2022, Version of Record 7 June 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.117641