Research On Default Risk Of Personal Credit Based On Factor Space Theory

In recent years,emerging Internet technologies such as artificial intelligence and mobile payment have gradually expanded to the traditional financial field in the form of Internet finance,forming a new type of personal credit background under the competition of personal credit business between traditional commercial banks and the Internet,and the factors of default risk are increasing.Therefore,based on the factor space theory,this paper studies the construction of reverse causal analysis and logistic regression algorithm,and applies them to the prediction of personal credit default risk:First,we make a more explicit mathematical description of the idea of reverse logic induction under the factor space theory,replace the background relationship with the background distribution,extract the reasoning rules with probability,and form the reverse causal analysis method.Using the method of reverse causality analysis,this paper studies the problem of credit card default risk identification,and puts forward the prediction model of reverse causality analysis of bank credit card default.The model takes the measured data of credit card as the background sample,extracts the inference rules,and verifies the prediction inference rules with the samples to be tested.Aiming at the unrecognized samples,a modified bank credit card default prediction model is proposed.Combined with KNN algorithm,the unrecognized samples caused by imbalanced data sets are identified twice to improve the recognition degree.The results on UCI data set show that the reverse causal analysis method is accurate and feasible in predicting credit card users' default.At the same time,the combination of reverse causal analysis and KNN algorithm is a good data mining method.Second,As the most representative online lending platform with a long history of personal credit development,P2 P can provide strong data support for exploring the default risk of personal credit,Logistic regression plays an important role in machine learning.At present,the research on logistic regression mainly stays in the application level.Therefore,based on the factor space theory,this paper further deepens the explanation of logistic regression,explores the explicit and implicit relationship of the underlying factors,and gives a reasonable expression of logistic regression from the perspective of explicit and implicit factors.Taking lending Club of the United States as an example,this paper selects the information data of borrowers in 2019 to establish a logistic regression prediction model of P2 P online loan credit default.Considering that the condition factors include multiple value States,the one hot idea is introduced to improve the accuracy of the algorithm.The accuracy rate,recall rate and other evaluation indexes were selected to compare and analyze the prediction effect of the model.The model results show that logistic regression can effectively predict the credit default risk of personal credit,and also provide a more in-depth explanation for the emergence of personal credit default risk under the new credit background.Finally,this paper applies the factor space theory to the field of credit default risk assessment of personal credit for the first time,and achieves good prediction results for the measured data of credit card and lending Club platform.This is not only of great significance for the in-depth study of the factor space theory,but also relatively enriches and improves the credit risk theory research system of personal credit,and provides certain reference value for credit institutions to realize the complementary advantages of traditional and network personal credit credit credit risk management and control and the optimization and innovation of personal credit business model.