Analysis of Individual Credit Scoring Based on Logistic Regression with the Adaptive Lasso

In order to analysis the influencing factors of personal credit scoring,regarding the personal credit risk as the research object.A bank personal credit data was used,and the Adaptive Lasso-Logistic regression model was adopted to analysis the influencing factors of personal credit risk from consumers.What is more,we compare this model with the traditional Logistic model and the Lasso-Logistic model.Regarding the proportion of correct classification results of ’good’ or ’bad’ consumers as the main measure,the result comparing with the traditional Logistic regression model and Lasso-Logistic regression model,shows that using the adaptive Lasso-Logistic regression model to establish the personal credit scoring model is superior to the traditional Logistic model and Lasso-Logistic model on variable selection and variable explanation and prediction accuracy.