Open AccessA novel method for credit scoring based on feature transformation and ensemble model
Author(s) -
Hongxiang Li,
Ao Feng,
Bin Lin,
Houcheng Su,
Zixi Liu,
Xuliang Duan,
Haibo Pu,
Yifei Wang
Publication year - 2021
Publication title -
peerj. computer science
Language(s) - English
Resource type - Journals
SCImago Journal Rank - 0.806
H-Index - 24
ISSN - 2376-5992
DOI - 10.7717/peerj-cs.579
Subject(s) - boosting (machine learning) , computer science , weighting , artificial intelligence , machine learning , transformation (genetics) , ensemble forecasting , feature engineering , ensemble learning , feature (linguistics) , process (computing) , data mining , pattern recognition (psychology) , deep learning , medicine , biochemistry , chemistry , linguistics , philosophy , radiology , gene , operating system
Credit scoring is a very critical task for banks and other financial institutions, and it has become an important evaluation metric to distinguish potential defaulting users. In this paper, we propose a credit score prediction method based on feature transformation and ensemble model, which is essentially a cascade approach. The feature transformation process consisting of boosting trees (BT) and auto-encoders (AE) is employed to replace manual feature engineering and to solve the data imbalance problem. For the classification process, this paper designs a heterogeneous ensemble model by weighting the factorization machine (FM) and deep neural networks (DNN), which can efficiently extract low-order intersections and high-order intersections. Comprehensive experiments were conducted on two standard datasets and the results demonstrate that the proposed approach outperforms existing credit scoring models in accuracy.