Risk Assessment And Analysis Of P2P Network Lending Platform Based On Classification Algorithm

P2P network lending platform is a combination of network lending and P2 P platform lending small-sum network lending way.It is also an important part of network finance.This new type of lending way supplies the traditional credit very well.Since its inception at the beginning of the 21 st century,it has developed rapidly on a global scale.With the appearance of PPDAI in 2007,China's P2 P network loan platform began to rise and get a rapid development from 2012.Since then,the number of network the growth rate of loan platforms is very fast.However,due to the initial oversight of the regulatory authorities,the platforms have many risk issues.These problems gradually exposed,including illegal fund-raising and so on.Therefore,the development speed of P2 P in our country has been slowed down and people have been doubted the P2 P network loan platform.The risks of P2 P loan platform includes two aspects: the environment of the platform and the behavior of lending and borrowing.Therefore,how to effectively identify the risks of P2 P loan platform has become the focus of the industry and academic research.There are many researches on P2 P platform at home and abroad,such as the research of platform's risk,government's supervision on platform,university students' lending,etc..Based on the data of net loan home,net loan eye and other net loan platforms and Lending Club loan data,this paper researched two aspects as the internal risk and external risk.Firstly,the open data of various online lending platforms are integrated and classified by cluster analysis.Using XGBoost and logistic regression,I have gotten more accurate expected results.Secondly,collecting borrower information for descriptive analysis from Lending Club to get accurate data analysis results.However,the risk prediction of P2 P loan platform needs to be further improved due to the difficulties of data' collection,the lack of open data and the reliability of data.This study shows that data mining can effectively Classification predict the risk of P2 P loan platform,in the meantime,which is helpful to reduce the risk of the platform and provide a healthy environment for the participants.The risk research of network lending platform is a realistic subject.Through multi-dimension and multi-method research,we can predict the platform's risk more effectively and provide a more healthy network lending platform.