AWS Certified Machine Learning – Specialty — Question 364

A company wants to use machine learning (ML) to improve its customer churn prediction model. The company stores data in an Amazon Redshift data warehouse.

A data science team wants to use Amazon Redshift machine learning (Amazon Redshift ML) to build a model and run predictions for new data directly within the data warehouse.

Which combination of steps should the company take to use Amazon Redshift ML to meet these requirements? (Choose three.)

Answer options

• A. Define the feature variables and target variable for the churn prediction model.
• B. Use the SOL EXPLAIN_MODEL function to run predictions.
• C. Write a CREATE MODEL SQL statement to create a model.
• D. Use Amazon Redshift Spectrum to train the model.
• E. Manually export the training data to Amazon S3.
• F. Use the SQL prediction function to run predictions.

Correct answer: A, C, F

Explanation

To train a model with Amazon Redshift ML, you must first define the input feature variables and the target variable from your data warehouse tables. Next, executing a CREATE MODEL SQL statement initiates the automated training process via Amazon SageMaker. Finally, once the model is trained, predictions are executed directly within Redshift queries using the SQL prediction function created during the model definition step.