Internet Credit Risk Scoring Based on Simulated Annealing and Genetic Algorithm

DOI

10.2991/ammsa-17.2017.84How to use a DOI?

Keywords

credit scoring; benchmark experiment; feature selection; simulated annealing; genetic algorithm

Abstract

Credit risk which can be reduced by credit scoring is the focus of financial risk control. In the construction of Internet credit scoring model, we encounter the problem that variables have high dimension. To solve this problem, feature selection is necessary. Simulated annealing and genetic algorithm can be used to do feature selection. This article gives an empirical analysis of individual credit valuation using data from a microfinance internet platform. Experimental result shows both of the logistic model based on simulated annealing and logistic model based on genetic algorithm have better prediction ability and model interpretability comparing to traditional full variable logistic regression. Also, the logistic model based on simulated annealing is slightly better than logistic model based on genetic algorithm

Copyright

© 2017, the Authors. Published by Atlantis Press.

Open Access

This is an open access article distributed under the CC BY-NC license (http://creativecommons.org/licenses/by-nc/4.0/).

Download article (PDF)

TY  - CONF
AU  - Ji Hu
AU  - Jiawen Cai
PY  - 2017/05
DA  - 2017/05
TI  - Internet Credit Risk Scoring Based on Simulated Annealing and Genetic Algorithm
BT  - Proceedings of the 2017 International Conference on Applied Mathematics, Modelling and Statistics Application (AMMSA 2017)
PB  - Atlantis Press
SP  - 373
EP  - 377
SN  - 1951-6851
UR  - https://doi.org/10.2991/ammsa-17.2017.84
DO  - 10.2991/ammsa-17.2017.84
ID  - Hu2017/05
ER  -