Classification Restricted Boltzmann Machine for comprehensible credit scoring model

作者：

Highlights：

• We propose a comprehensible model for credit risk assessment using a scoring table.

• We use Restricted Boltzmann Machine to determine scoring points in a scoring table.

• We deal with the imbalanced data by applying the geometric mean criterion.

• The quality of the presented method is evaluated on four credit scoring datasets.

摘要

•We propose a comprehensible model for credit risk assessment using a scoring table.•We use Restricted Boltzmann Machine to determine scoring points in a scoring table.•We deal with the imbalanced data by applying the geometric mean criterion.•The quality of the presented method is evaluated on four credit scoring datasets.

论文关键词：Credit scoring,Comprehensible model,Restricted Boltzmann Machine,Imbalanced data

论文评审过程：Available online 18 October 2014.

论文官网地址：https://doi.org/10.1016/j.eswa.2014.10.016