As a Machine Learning Engineer working on improving a customer churn prediction model in AWS, you need to handle missing data that could impact your model's performance. Which of the following techniques is the most appropriate for managing missing data in your dataset using AWS services?