摘要
大数据在供应链金融中的应用,推动了供应链信用评估模式和风险管理方法的重大变革。通过将前沿大数据征信技术与评分卡方法相结合,以整车制造供应链为场景,对供应链金融信用风险进行测度。首先,借助Python软件,从“企查查”API数据接口和Wind数据库获取相关数据,对27家核心企业122条供应链多维指标进行数据挖掘、WOE编码和变量筛选,构建指标体系。然后,运用大数据和人工智能建模思路,建立涵盖14个特征解释变量的Logistic回归模型,并运用多种工具训练和改进模型,形成可应用于实务的Logistic评分卡。经实证检验,最终确定的信用评估模型区分能力强,风险预测准确率能达到96.77%。基于大数据的Logistic评分卡将供应链信用等级数字化,相较于传统的信用评级更具有实用性,因此大数据技术的运用对提升供应链金融信用风险评估和管理水平具有重要价值。
The application of big data in supply chain finance has driven a major change in the supply chain credit assessment mode and risk management methodology.By combining the cutting-edge big data credit technology with the scorecard method,we measure the credit risk of supply chain finance by taking the whole vehicle manufacturing supply chains as the scenario.Firstly,with the help of Python,relevant data were obtained from the“Qichacha”API and Wind database,data mining,WOE coding and variable screening were carried out for 122 supply chain multidimensional indicators of 27 core enterprises to construct an indicator system.Then,using big data and AI modeling technologies,a logistic regression model covering 14 characteristic explanatory variables is established,and multiple tools are used to train and improve the model to form a logistic scorecard that can be used in practice.After empirical testing,the finalized credit assessment model has strong differentiation ability and can reach 96.77%risk prediction accuracy.The logistic scorecard based on big data digitizes the supply chain credit rating,which is more practical compared to the traditional credit rating,so the use of big data is of great value to improve the credit risk assessment and management of supply chain finance.
作者
周雷
赖姝牟
付煜棋
ZHOU Lei;LAI Shumou;FU Yuqi(School of Business,Suzhou Vocational University,Suzhou,Jiangsu 215104,China;School of Cornputer Engineering,Suzhou Vocational University,Suzhou,Jiangsu 215104,China)
出处
《金融教育研究》
2023年第1期64-73,共10页
Research of Finance and Education
基金
教育部人文社会科学研究基金项目“新时代大学生互联网金融风险认知、风险偏好与投资行为研究”(19YJCZH272)
江苏高校哲学社会科学研究项目“数字经济时代金融科技服务实体经济高质量发展研究”(2022SJYB1630)
苏州市职业大学校级研究性课程项目“数智化时代大数据征信应用研究”(SZDYKC-220201)。
