Predicting the Failure of Chinas Online Loan Platform with Machine Learning Models

• Main Authors: HUANG, JHIH-HUEI, 黃智暉
• Other Authors: YE, JHEN-YIN
• Format: Others
• Language: zh-TW
• Published: 2019
• Online Access: http://ndltd.ncl.edu.tw/handle/5m43g4

碩士 === 國立屏東大學 === 商業自動化與管理學系碩士班 === 107 === 　　In recent years, the P2P(Peer-to-Peer) online lending platform has developed rapidly. China is currently the worlds largest P2P lending market. But in the past few years, thousands of Chinese P2P platforms have failed and have closed down. Most P2P online lending platforms focus on the research of platform mechanisms and transaction data. This study takes Chinas P2P online lending platform as the research object, collects the public information of about 3,033 online lending platforms from Chinas P2P online lending platform from 2014 to 2018 as research variables, and constructs predictive impact platform persistence mode by four machine learning methods. The results of the study show that all four machine learning models have excellent or excellent predictive power. The research results also found that company licenses, bank depository, and supervision characteristics are important variables that affect the survival of the platform. The results of this study can be used as a support tool for investment strategy evaluation by borrowers and investors of the P2P online lending platform.