The trading amount of P2P network lending is rising, and the research of P2P trading data receives much attention. The factor analysis of the success rate of network loan is one of the important research topics. The previous papers on this issue mainly adopt multi-linear regression method, ignoring the problem of multi-collinearity between the variables and the finding of "optimal" regression model. This paper uses the Lasso regression method to establish the regression model with optimal subset of variables, which can analyze the factors that affect the success rate of network borrowing, avoiding the multi-collinearity of the model interference and improving the prediction accuracy of the model. This paper empirically analyzes the borrowing and lending data from the Lending Club platform, and the result shows that our method is significantly superior to the compared approach in the aspects of fitting precision of the model and avoiding the multi-collinearity.