This project is a solution to the KKBox Churn Prediction Challenge, where the objective is to predict whether a user will churn by ending their current subscription.

KKBOX is Asia's leading music streaming service, with over 30 million tracks in its Asia-Pop music library. KKBOX offers a generous, unlimited version of its service to millions of users, supported by advertising and paid subscriptions. The business model of KKBOX depends heavily on accurately predicting churn of its paid users.

In this project, we leverage machine learning techniques to develop a churn prediction model that can accurately forecast which users are likely to churn, enabling KKBOX to take proactive measures to retain its customers.

We utilize various feature engineering techniques, and a variety of machine learning models, such as decision trees, random forests, and gradient boosting machines, to build and optimize our churn prediction model. The ultimate goal of this project is to develop an accurate and reliable churn prediction model that can be deployed in KKBOX's business operations to help improve customer retention and reduce churn.