The invention discloses a mobile banking marketing customer screening method fusing multiple machine learning models, relates to the technical field of financial data screening processing, and solvesthe technical problem of insufficient precision degree of a data processing method for mobile banking activity promotion precision marketing in the existing banking industry, and the mobile banking marketing customer screening method comprises the following steps of S1, carrying out data preparation; s2, generating an active label; s3, enabling the data to meet the requirements of entering the model; s4, performing training, testing and predicting by adopting an active customer identification model and a marketing response customer identification model; s5, calculating the matching degree of the customer asset grade score and the marketing activity; s6, calculating a customer marketing value in combination with the active activity probability, the marketing response probability, the assetlevel and the marketing activity matching degree; s7, sorting the clients according to the marketing values of the clients from high to low, and screening out the clients with high marketing values. Customer data with high marketing value can be accurately screened out.