基于图神经网络的企业信用评级方法
A Method for Corporate Credit Rating Based on Graph Neural Network
摘要: 企业作为市场经济的主要参与者，其信用风险一直以来备受学术界和金融界的关注。近年来，图神经网络模型被广泛应用于金融问题的研究，在这些工作中，企业被视作顶点，企业间的关系被视作边，从而利用模型完成企业的信用评估。上述这些方法往往将企业内部各指标一视同仁，没有分析各数据间的内在关系。基于前人经验，我们筛选出29项企业财务数据指标，将各指标抽象为顶点，深入分析各指标之间的关系，构建指标相似度矩阵，在此基础上利用最小生成树算法实现企业的图映射，并搭建图神经网络模型，得到其嵌入表示，最后结合神经网络分类器完成预测任务。应用真实企业数据的实验结果表明，模型能较好的完成企业的多层级信用水平评级。
Abstract: As a major participant in the market economy, the credit risk of enterprises has always attracted the attention of academic and financial circles. In recent years, the graph neural network model has been widely used in the research of financial issues. In these works, enterprises are regarded as vertices, and the relationship between enterprises are regarded as edge, so that the model can be used to complete the credit evaluation of enterprises. The above-mentioned methods often treat all indicators within the enterprise equally, and do not analyze the internal relationship between var-ious data. Based on previous experience, we screened out 29 corporate financial data indicators, abstracted each indicator as a vertex, deeply analyzed the relationship between indicators, and constructed an indicator similarity matrix. On this basis, we used the minimum spanning tree algo-rithm to realize the graph of the enterprise mapping, and build a graph neural network model, get its embedded representation, and finally combine the neural network classifier to complete the prediction task. The experimental results using real enterprise data show that the model can better complete the multi-level credit rating of enterprises.
文章引用：张晓薇, 胡越. 基于图神经网络的企业信用评级方法[J]. 应用数学进展, 2023, 12(4): 1838-1847. https://doi.org/10.12677/AAM.2023.124190

参考文献

[1] 孟杰. 企业信用风险评估方法综述[J]. 中国国际财经(中英文), 2018(2): 290. [Google Scholar] [CrossRef] 
[2] Altman, E.I. (1968) Financial Ratios, Discriminant Analysis and the Prediction of Corporate Bankruptcy. Journal of Finance, 23, 193-194. [Google Scholar] [CrossRef] 
[3] 梁世栋, 郭仌, 李勇, 方兆本. 信用风险模型比较分析[J]. 中国管理科学, 2002, 10(1): 18-23.
[4] Golbayani, P., Florescu, I. and Chatterjee, R. (2020) A Comparative Study of Forecasting Corporate Credit Ratings Using Neural Networks, Support Vector Machines, and Decision Trees. The North American Journal of Economics and Finance, 54, Article ID: 101251. [Google Scholar] [CrossRef] 
[5] 杨保安, 季海, 徐晶, 温金祥. BP神经网络在企业财务危机预警之应用[J]. 预测, 2001, 20(2): 49-54+68.
[6] Cheng, D., Niu, Z. and Zhang, Y. (2020) Contagious Chain Risk Rating for Networked-Guarantee Loans. Proceedings of the 26th ACM SIGKDD International Conference on Knowledge Discovery & Data Mining, Virtual Event, 6-10 July 2020, 2715-2723. [Google Scholar] [CrossRef] 
[7] Feng, B.J., Xu, H.N., Xue, W.F. and Xue, B.D. (2020) Every Corporation Owns Its Structure: Corporate Credit Ratings via Graph Neural Networks. ArXiv: 2012.01933v1.
[8] 陈丽君. 互联网供应链融资小微企业信用风险评估体系研究——基于电商平台融资模式[J]. 商业会计, 2018(20): 76-77.
[9] 池仁勇, 张帆. 软信息与硬信息孰轻孰重——中小企业授信与信用风险视角[J]. 华东经济管理, 2020, 34(3): 112-118.
[10] 段翀. 基于PFM模型与倾向匹配得分法的中小企业信用风险评估[J]. 征信, 2020, 38(6): 35-42.
[11] Wang, X., Ye, Y. and Gupta, A. (2018) Zero-Shot Recognition via Semantic Embeddings and Knowledge Graphs. 2018 IEEE/CVF Conference on Computer Vision and Pattern Recognition, Salt Lake City, 18-23 June 2018, 6857-6866. [Google Scholar] [CrossRef] 
[12] Zhang, M. and Chen, Y. (2018) Link Prediction Based on Graph Neural Networks. Proceedings of the 32nd International Conference on Neural Information Processing Systems, Mont-réal, 3-8 December 2018, 5171-5181.
[13] Bojchevski, A. and Günnemann, S. (2018) Deep Gaussian Embedding of Graphs: Unsupervised Inductive Learning via Ranking. 6th International Conference on Learning Representations, ICLR 2018, Vancouver, 30 April-3 May 2018, 1-13.
[14] Zitnik, M., Agrawal, M. and Leskovec, J. (2018) Modeling Polypharmacy Side Effects with Graph Convolutional Networks. Bioinformatics, 34, i457-i466. [Google Scholar] [CrossRef] [PubMed]
[15] Liu, J., Lu, Z. and Du, W. (2019) Combining Enterprise Knowledge Graph and News Sentiment Analysis for Stock Price Prediction. Proceedings of the 52nd Hawaii Interna-tional Conference on System Sciences, Wailea, 8-11 January 2019, 1247-1255. [Google Scholar] [CrossRef] 

为你推荐