Deep neural networks with L1 and L2 regularization for high dimensional corporate credit risk prediction

作者：

Highlights：

• This study adds more external information to predict corporate credit risk (CCR).

• This study proposes the HDNN algorithm to predict high dimensional CCR.

• This study proves that adding L2 constraints on a single L1 regularization can prevent L1 from failing in DNN algorithms.

• The performance of the HDNN algorithm is better than that of competing algorithms.

摘要

•This study adds more external information to predict corporate credit risk (CCR).•This study proposes the HDNN algorithm to predict high dimensional CCR.•This study proves that adding L2 constraints on a single L1 regularization can prevent L1 from failing in DNN algorithms.•The performance of the HDNN algorithm is better than that of competing algorithms.

论文关键词：High dimensional data,Credit risk,Deep neural network,Prediction,L1 regularization

论文评审过程：Received 29 December 2021, Revised 8 September 2022, Accepted 17 September 2022, Available online 21 September 2022, Version of Record 30 September 2022.

论文官网地址：https://doi.org/10.1016/j.eswa.2022.118873