Font Size: a A A

Application Of Data Mining Combination Model In Personal Credit Risk Assessment

Posted on:2018-05-25Degree:MasterType:Thesis
Country:ChinaCandidate:R ZhangFull Text:PDF
GTID:2359330518950311Subject:Applied Statistics
Abstract/Summary:PDF Full Text Request
With the rapid development of personal consumption credit,personal credit card is showing an increasing trend.In order to effectively carry out personal credit assessment and risk forecasting,more and more business departments need to use scientific and efficient data mining tools as auxiliary tools to go the actual business.Based on Lending Club's personal consumption credit data,this paper applies data mining technology to P2 P credit field.After the relevant theoretical analysis,this paper has done a lot of basic work,such as data acquisition,data conversion,data integration and data cleanup,from the data source to ensure the correctness,consistency,integrity and reliability of the data.The paper then carries out the user's portrait of the borrower,from the descriptive statistics first grasp the borrower's relevant information.Then,based on the standardized data frame after standardization,a single prediction model is established,including logistic regression model and artificial neural network model,which are optimized in the process of model establishment to achieve the best single model prediction results.Then,the effective predictor of the borrower is selected by the training of the random forest,the interference of the redundant index to the prediction result is eliminated,and the output result of the logistic regression(ie,the probability value)is taken as the important variable,which is used as artificial neural network's input variables of the model establish a combined model.It is proved that the combined model is more accurate than the single model,the stability is stronger and the explanation can be better.Finally,this paper analyzes some potential problems and analyzes the follow-up work.The ultimate goal of this paper is to build an effective risk assessment model by using the borrower's portrait model which is already mature in the foreign P2 P network loan mode,to promote the construction of the domestic personal credit risk assessment system,to speed up the commercial practice of personal consumption credit Provide some technical support.
Keywords/Search Tags:data mining, credit assessment, random forest, Logit model, artificial neural network
PDF Full Text Request
Related items