Investment Decision Analysis Of P2P Lending Based On Support Vector Machine

P2P(Peer to Peer)lending is an internet-based lending method for person to person.This simple and convenient network lending model not only quickly solves the borrower's shortage of funds,but also brings new financial management to the majority of investors.Chinese P2 P lending is faulty in law supervision,individual credit rating system,and the phenomenon of default often occurs on platform.It greatly hurts the interests of the platform and the lender.Therefore,it is of great practical significance to establish a credit risk assessment model,help the platform to carry out risk supervision,and assist the lender to make appropriate investment strategies.In this paper,support vector machine algorithm,the kernel regression estimation and portfolio optimization model are used to help lenders distinguish the loan risk and allocate the investment amount reasonably.The main contents are as follows:Firstly,based on support vector machine,an investment decision model is constructed.The specific contents are as follows:(1)The support vector machine is used to classify the loans' status,and a classification hyperplane is obtained.Then calculate the probability of default for each loan based on the distance from the loan to the classification hyperplane.(2)Use the probability of default to describe the degree of similarity between loans,and estimate the return and risk by the degree of similarity and kernel regression.(3)Combined with the return and risk,we use the portfolio theory to study the income and risk of the loan portfolio,and obtain the optimal loan portfolio.Secondly,we analyze the factors that affect the rate of return in P2 P lending platform.From the perspective of the platform,100 platforms are selected as the samples to analyze the factors that affect the expected rate of return.The results show that the main asset type for P2 P lending platform and the loan period have a significant impact on the expected rate of return for platform.Finally,based on support vector machine,we select the data of Renrendai platform to verify the investment model.From the perspective of the lenders,the validity of the investment decision model based on support vector machine is verified by the data of credit certification object.The results show that the proposed model is superior to the logistic regression model in terms of classification accuracy,yield forecast accuracy,and portfolio yield forecast accuracy.