文献类型：期刊文献

英文题名：Early Warning for Internet Finance Industry Risk: An Empirical Investigation of the P2P Companies in the Coastal Regions of China

作者：Ma, Liyi[1];Wang, Yane[1];Ren, Chengmei[1];Li, Hui[1];Li, Yingxia[1]

第一作者：马丽仪

通讯作者：Ma, LY[1]

机构：[1]Beijing Union Univ, Sch Management, Beijing 100010, Peoples R China

第一机构：北京联合大学管理学院

通讯机构：[1]corresponding author), Beijing Union Univ, Sch Management, Beijing 100010, Peoples R China.|[1141755]北京联合大学管理学院;[11417]北京联合大学;

年份：2020

卷号：106

期号：sp1

起止页码：295-299

外文期刊名：JOURNAL OF COASTAL RESEARCH

收录：;Scopus(收录号：2-s2.0-85088707562);WOS:【SSCI(收录号:WOS:000550684000069)，SCI-EXPANDED(收录号:WOS:000550684000069)】；

基金：The authors thank for the supports from the Humanities and Social Sciences Fund Project of the Ministry of Education (Research on the Risk Evolution and Control System of High-tech Industry Cluster from the Perspective of Cluster Network Structure, 15YJCZH114), Beijing Social Science Foundation Project (Research on the Cultivation of Core Information Technology Industry in Beijing, 18GLB028), and Premium Funding Project for Academic Human Resources Development in Beijing Union University (BPHR2017CS17, SK50201903).

语种：英文

外文关键词：Lending platforms; risk warning; decision tree model

摘要：Recent years have seen the rapid development of Internet finance in China, and various peer-to-peer (P2P) lending platforms have been released. However, the emergence of diverse problems, such as absconding with the money and fraudulent platforms, has exposed the lack of supervision of P2P lending platforms. This paper collected 26 dimensions of data from 1292 P2P companies in the coastal regions of China from third-party websites. Risk warning and machine learning were incorporated to construct a risk warning model for P2P lending platforms. This model consists of a decision tree model, naive Bayes model, and support vector machine model. Correlation analysis was used to analyze the data source, which is the data predicted by the above models, and the target variables. Type I error rate was utilized to determine the optimal model. The conclusions are as follows: The indicators of registered capital, changes of business scope, and platform background have a great impact on the accuracy of P2P risk prediction; the decision tree model is capable of predicting problematic platforms; and the decision tree model, support vector machine model, and naive Bayes model have a descending order in accuracy. Hence, the decision tree model is the best model for risk warning.