Research On Personal Finance Credit Risk Assessment Model Of Internet Finance Based On PSO-SVM

In recent years,the Chinese economy is developing at an alarming rate.At the same time as the overall consumption power of the nation rises,the Chinese people's awareness of credit consumption is constantly awakening.Since 2006,China's Internet finance industry is developing rapidly.According to the data of the Network Lending Eye Research Institute,as of December 31,2018,the number of P2 P online lending platforms in China totaled 6,591,including 1,609 operating platforms,and 4,982 outages were rectified due to various problems.Part of the P2 P platform that was out of service was due to its own internal operation problems.Among them,some platforms are caused by the failure of the borrower's credit risk management to cause the platform's own capital chain to break..There are many differences between those who borrow from the Internet financial platform and the borrowers who apply for loans to traditional banks.Most of the consumers who borrow from the Internet financial platform have problems such as rushing money and difficulties in traditional bank loans.However,due to the fact that most borrowers do not need to provide collateral when borrowing from the Internet financial platform.From this perspective,the risk of default by borrowers facing the Internet financial platform is much greater than that faced by traditional banks.From the perspective of the Internet financial platform itself,the platform needs to establish its own set of evaluation systems and assessment methods to conduct risk assessments on borrowers to reduce the risk of bad debts caused by misjudging the credit status of borrowers.The study of personal credit assessment models is an urgent task.Considering the current situation of personal credit business in the Internet financial industry and the difficulties faced by the Internet financial platform,this paper proposes to establish an Internet financial personal credit evaluation model based on the support vector machine.For the reason that the support vector machine cannot eliminate the evaluation error caused by the redundancy of the input vector index,this paper uses the principal component analysis method to screen the input vector indicators.In this paper,the particle swarm optimization algorithm with fast optimization characteristics is proposed to optimize its parameters,because the accuracy of the support vector machine classification model mainly depends on the selection of kernel function parameters.Based on the above foundation,this paper has implemented the support vector machine model before and after optimization,and further verified the support vector machine model before and after particle swarm optimization with real internet financial data.In view of the difference between the first type of misjudgment and the second type of misjudgment in the credit evaluation process,when comparing the performance of the model,this paper,not only compares the accuracy of the model,but also increases the comparison between the first type of false positive rate and the second type of false positive rate.By comparing the support vector machine models before and after particle swarm optimization,the support vector machine model optimized by particle swarm optimization algorithm has better evaluation effect than the unoptimized support vector machine,and the classification accuracy and stability of the model have been significantly improved.The research shows that the particle swarm optimization algorithm has good performance in optimizing the support vector machine personal credit risk assessment model.It is of great practical significance to apply this model to the field of personal credit assessment.