Amazon Sagemaker Modeling

Question 1

Which of the following models are supervised algorithms?
A. Clustering

B. Classification

C. Association rule mining

D. Regression

Answer 1

B. Classification

D. Regression

Question 2

You would like to turn your Amazon SageMaker machine learning models and endpoints into customer-facing applications. You decide to put these on a single web server that can be accessed by customers via a browser. However, you realize that the web server is not inherently scalable; if it receives a lot of traffic, it could run out of CPU or memory. How can you make this approach more scalable and secure?

Answer 2

A. Create an IAM role, so the webserver can access the SageMaker endpoints.

B. Deploy a load balancer and set up autoscaling.

D. Keep the operating system and language runtimes for the web server patch secured.

Question 3

For the preceding situation, what would be a better AWS service to automate server and operating system maintenance, capacity provisioning, and automatic scaling?
A. AWS Lambda

B. AWS Fargate

C. AWS ELB

Answer 3

A. AWS Lambda

Question 4

Amazon SageMaker is a fully managed service that enables you to quickly and easily integrate machine learning-based models into your applications. It also provides services such as notebook, training, and endpoint instances to help you get the job done.
A. TRUE

B. FALSE

Answer 4

A. TRUE

Question 5

Chose three correct statements from the following:
A. Notebook instances clean and understand data.

B. Training instances use data to train the model.

C. Endpoint instances use models to produce inferences.

D. Notebook instances clean, understand, and build models.

E. Training instances are used to predict results.

Answer 5

A. Notebook instances clean and understand data.

B. Training instances use data to train the model.

C. Endpoint instances use models to produce inferences.

Question 6

What is the first step of creating a notebook?
A. Give it a name.

B. Choose a kernel.

C. Starting developing code in paragraph format.

Answer 6

B. Choose a kernel.

Question 7

Linear learner and XGBoost algorithms can be used in supervised learning models such as regression and classification.
A. TRUE

B. FALSE

Answer 7

A. TRUE

Question 8

Which of these statements about hyperparameter tuning is true?
A. Hyperparameter tuning is a guaranteed way to improve your model.

B. Hyperparameter tuning does not require any input values.

C. Hyperparameter tuning uses regression to choose the best value to test.

D. Hyperparameter tuning is an unsupervised machine learning regression problem.

Answer 8

C. Hyperparameter tuning uses regression to choose the best value to test.