Yuan Yan
Guangdong Mechanical & Electrical Polytechnic, School of Economics and Trade, Guangzhou 510515, China.
Comput Intell Neurosci. 2022 Jul 20;2022:9274737. doi: 10.1155/2022/9274737. eCollection 2022.
In the free flow of financial factors oriented to capital, returns will be accompanied by the concentration and diffusion of financial resources to form regional financial spatial differences, which is an objective phenomenon of regional financial practice. Localized regional financial risks may appear in the process of regional financial practice in each region. To address the abovementioned problems, we propose a model for regional financial risk analysis based on the DCN deep learning model. The main contents are as follows: elaborating the financial risk transmission mechanism involving intra- and interregional financial risks, sorting out the relationship between sectors as clues; the designing process of regional financial risk index as well as the measurement method, and the regional financial risk index for typical regions is measured and found to be at peak in 2017 with a risk index of 0.58; and the construction of an early warning model based on the value of the regional financial risk index and the expansion of the RNN network applied to the construction of the regional financial risk early warning system. Based on the construction of the RNN network application risk early warning system, the three types of risks, payment risk, loan loss risk, and market risk with the percentages of 49.62%, 26.82%, and 23.56%, respectively, are derived, and the focus is on their supervision and management in the follow-up work.
在以资本为导向的金融要素自由流动中，收益将伴随着金融资源的集中与扩散，从而形成区域金融空间差异，这是区域金融实践中的一种客观现象。在各地区的区域金融实践过程中，可能会出现局部性的区域金融风险。为解决上述问题，我们提出一种基于深度卷积网络（DCN）深度学习模型的区域金融风险分析模型。主要内容如下：阐述涉及区域内和区域间金融风险的金融风险传导机制，以梳理部门间关系为线索；区域金融风险指数的设计过程以及测量方法，对典型地区的区域金融风险指数进行测量，发现其在2017年达到峰值，风险指数为0.58；基于区域金融风险指数的值构建预警模型，并将循环神经网络（RNN）进行扩展应用于区域金融风险预警系统的构建。基于RNN网络应用风险预警系统的构建，得出支付风险、贷款损失风险和市场风险这三种风险类型，其占比分别为49.62%、26.82%和23.56%，后续工作将重点关注对它们的监督和管理。
