Graph convolutional network-based credit default prediction utilizing three types of virtual distances among borrowers

作者：

Highlights：

• Combined graph convolutional network-based default prediction model is proposed.

• Individual GCNs with different types of inputs with different weights are combined.

• The contribution of borrower‘s features to predict default is measured in the GCN.

• Applications to Lending Club data show that our model outperforms existing models.

摘要

•Combined graph convolutional network-based default prediction model is proposed.•Individual GCNs with different types of inputs with different weights are combined.•The contribution of borrower‘s features to predict default is measured in the GCN.•Applications to Lending Club data show that our model outperforms existing models.

论文关键词：Credit scoring,Peer-to-peer lending,Deep learning,Graph convolutional network

论文评审过程：Received 29 July 2020, Revised 16 November 2020, Accepted 28 November 2020, Available online 10 December 2020, Version of Record 15 December 2020.

论文官网地址：https://doi.org/10.1016/j.eswa.2020.114411