基于机器学习算法的企业债券违约预警模型  

A Machine Learning Based Enterprise Bond Default Warning Model

作　　者：胡传胜 周志国 HU Chuansheng;ZHOU Zhiguo(Anhui Jiyuan Software Co.,LTD.,Hefei 230088,China)

机构地区：[1]安徽继远软件有限公司,安徽合肥230088

出　　处：《长江信息通信》2024年第12期173-175,202,共4页Changjiang Information & Communications

摘　　要：本研究旨在通过大数据方法构建企业债券违约预警模型,利用逻辑回归、XGboost、支持向量机等机器学习算法,结合企业用电、财务等多维数据,搭建企业违约预警指标体系,训练并预测当期企业债券违约概率,划分企业风险等级,降低投资风险。This study aims to construct a corporate bond default warning model using big data methods,utilizing machine learning algorithms such as logistic regression,XGboost,and support vector machines,combined with multidimensional data such as enterprise electricity and finance,to establish a corporate default warning indicator system,train and predict the default probability of current corporate bonds,classify enterprise risk levels,and reduce investment risks.

关 键 词：债券违约 投资风险 逻辑回归 XGBoost算法 支持向量机 

分 类 号：TP181[自动化与计算机技术—控制理论与控制工程]