Customer churn has a significant impact on enterprises as it directly affects their revenue and profit. This paper focuses on analyzing user behavior data based on a machine learning-based user churn prediction model, constructs a prediction model, and verifies its effectiveness. This aims to help enterprises better understand user behavior and prevent user churn in advance. Meanwhile, by collecting and analyzing user data, multiple machine learning algorithms are used for modeling and evaluation to determine the optimal prediction model, which is then applied to practical business scenarios. The experimental results show that the model can accurately predict user churn, providing strong support for enterprises to develop personalized marketing strategies and enhancing user retention rates and enterprise competitiveness.