Precise intelligent user classification is a critical foundation for modern commercial banks to achieve digital transformation in the era of big data. This article presents a comprehensive model for classifying credit card users based on whether they opt for installment payments. Using data sourced from monthly comprehensive data provided by the Jilin Bank Credit Card Center and after performing essential operations like dataset construction and data cleaning, the article employs the NSGA-II algorithm for feature selection to determine the feature set to be used in the model. The LightGBM algorithm is then employed to train on millions of customer data records based on the final feature set, producing user classification results. Real-world application results demonstrate that within 3 months, the model proposed in this article increased the installment application rate by 3.76% and improved the outbound call success rate by 1.16%.