Abstract: While P2P is experiencing explosive growth, it also faces significant credit risks. The prediction of borrower default risk is an important method to reduce credit risk. Taking the data taken on the platform of "Everyone's Loan" as the research sample, the credit evaluation index system of borrowers is constructed, and the dual-logistic regression model is used to establish the credit risk assessment model of borrowers. The results show that the borrowing period, the borrower's age, credit rating and overdue times have the most significant impact on the borrower's credit risk, followed by education, successful borrowings, borrowing rates, and real estate.

Key words: P2P, credit risk assessment, dual-logistic regression model