In recognizing the significance of retaining current consumers to thrive in this competitive landscape, this paper aims to predict customers' churn probability based on the background and behaviors of previous customers. The paper utilizes churn probability as a proactive means to identify customers at a high risk of leaving, serving as a reference for Telco to make informed decisions and take actions aimed at enhancing customer loyalty. This paper pre-target customers who have a high risk of leaving based on our churn probability with at least over 80% accuracy rate. And then use the churn probability as the reference to help Telco make better decisions and take actions to retain customers, such as providing a private discount or coupon. This paper limits the choice of customers in the services provided by telecom companies. However, customer attrition decisions may also be related to the quality of the service and the level of price, which the study was unable to quantify. The study addresses two primary challenges: determining models that accurately predict customer churn and identifying the characteristics of customers more prone to churn.