Churn prediction is one of the most difficult Big Data use cases. It is the most important indicator for a robust and expanding company, regardless of size or sales channel. Consumer churn, defined as a consumer leaving an established relationship with a company, is an important topic that has been extensively researched for both academic and commercial purposes. When a company's clients discontinue doing business with it, this is referred to as churn. In online commerce, a customer is considered churned when his or her transactions are outdated for more than a certain period. When a customer churns, the company suffers a loss that includes not just the lost revenue from the lost customer, but also the expenditures of further marketing to acquire new customers. The key goal of every online business is to reduce client churn. Customer attrition is detrimental to a company's bottom line. As a result, accurate customer churn prediction is critical for organizations seeking to improve client retention and corporate profits. However, there are challenges with assessing client attrition using standard methodologies in the case of mobile commerce. In this article, a Machine Learning (ML) model for predicting churn in mobile commerce is being built. By using the customer dataset, three techniques such as Support Vector Machine (SVM), Linear Discriminant Analysis (LDA), and Naive Bayes (NB) are used to estimate churn. These three algorithms are compared using accuracy metrics in a study. When compared to other methodologies, LDA had a better churn prediction accuracy of 92.53%.