AWS Certified Data Analytics Specialty 2021 – Hands-On - Amazon ML Service

The video tutorial provides an overview of Amazon Machine Learning (ML) service, highlighting its features, such as visualization tools and wizards for model creation without coding. It explains how Amazon ML can be used for predictions, both in real-time and batch processing, and discusses its fully managed nature. The tutorial also covers the limitations of Amazon ML, including its outdated status and algorithm restrictions. It provides examples of suitable applications like fraud detection and forecasting, and outlines the cost model and data limitations. Finally, it identifies anti-patterns and tasks unsuitable for Amazon ML.

What is a key feature of Amazon ML that simplifies model creation?

• It requires manual server provisioning.

• It requires extensive coding knowledge.

• It provides visualization tools and wizards.

• It only works with on-premise data.

Which of the following is a limitation of Amazon ML?

• It can handle terabytes of data by default.

• It has limited algorithm choices.

• It is not a fully managed service.

• It supports deep learning algorithms.

For which task is Amazon ML particularly suitable?

• Unsupervised clustering

• Fraud detection

• Sequence prediction

• Deep learning

How does Amazon ML charge users?

• Flat monthly fee

• Annual subscription

• Based on the number of users

• Pay for what you use

What is the maximum default training data size for Amazon ML?

• 10 TB

• 1 TB

• 100 GB

• 10 GB

Which of the following is NOT an ideal use case for Amazon ML?

• Product demand forecasting

• Deep learning

• Binary classification

• Regression tasks

What should you do if you need to run more than five simultaneous jobs on Amazon ML?

• Upgrade your subscription

• File a support ticket

• Use a different service

• Increase your data limit