Font Size: a A A

Research On Risk Evaluation Of Peer-to-peer Lending Platforms Based On Data Mining Technology

Posted on:2019-02-23Degree:MasterType:Thesis
Country:ChinaCandidate:J R KangFull Text:PDF
GTID:2429330566458723Subject:Management Science and Engineering
Abstract/Summary:PDF Full Text Request
With the development of e-commerce and network mobile technology,the time and geographical limitation of the traditional financial industry is broken,the Internet technology is prevailing,and has undergone great changes.P2 P lending is a single member of the direct lending,financial supply and demand information directly posted online and matching,financial "de media" has become a major trend.P2 P lending as a new form of financing,has following advantages:(1)make up for the shortcomings of the financial institutions,(2)borrowers can rely on P2 P platform to meet their own financing needs,(3)investors can also get higher interest income.Data mining draws on the ideas of artificial intelligence,machine learning,statistics,can explore its potential information in large data,reduce information asymmetry and moral hazard.The characteristic of P2 P platform determines the risk identification and platform classification of P2 P platform using data mining technology,which has certain feasibility and practical significance.This paper collects the real data of P2 P platform,using DP algorithm and the C5.0 algorithm,analyzes the financial attribute and Internet attribute of P2 P platform,classifies and categorizes P2 P platform in our country,main contents are as follows:(1)Aiming at the imperfect and objectivity of the risk evaluation index of the existing P2 P platform,the paper puts forward the risk identification model of P2 P platform with financial and Internet attributes.(2)Select a variety of clustering algorithms for P2 P platform risk identification,realize DP and P2 P platform data seamless convergence.(3)In order to obtain the optimal risk recognition effect of P2 P platform,a new risk identification is constructed,which combines DP clustering with multiple representative point clustering effectiveness index CDbw.(4)Using C5.0 decision tree to train classifier and study classification rules,analyze P2 P platform data from multiple dimensions,and provide practical reference for investors to choose the platform for investment.The results of paper show that the risk identification index of P2 P platform and the application of DP and C5.0 decision tree can effectively cluster and classify P2 P platform.It helps investors to identify the risk of P2 P platform,grasp the risk level,evade the financial risk,and help investors make rational investment decisions.
Keywords/Search Tags:Peer-to-peer lending platforms, Information disclosure, Density peak clustering algorithm, Risk evaluation
PDF Full Text Request
Related items