A retail company is developing a customer churn prediction model on AWS to identify customers likely to cancel their subscriptions. The training dataset includes customer purchase history, support interaction logs, and subscription data. The purchase history and support logs are stored in Amazon S3, while the subscription data resides in an on-premises PostgreSQL database. The dataset has two major challenges: A class imbalance where very few customers are labeled as "churned", impacting the model’s learning. There are strong feature interdependencies among categorical features (e.g., "membership tier") and numerical features (e.g., "purchase frequency"). The ML engineer needs to select an Amazon SageMaker built-in algorithm to train the model and address the challenges with the least operational effort. Which algorithm should the ML engineer use to meet these requirements?