Microsoft Microsoft Certifications DP-100 Questions & Answers

• Question 181:

  You have the following Azure subscriptions and Azure Machine Learning service workspaces:

  

  You need to obtain a reference to the ml-project workspace. Solution: Run the following Python code:

  

  Does the solution meet the goal?

  A. Yes

  B. No

  Correct Answer: B

• Question 182:

  You create a training pipeline by using the Azure Machine Learning designer. You need to load data into a machine learning pipeline by using the Import Data component.

  Which two data sources could you use? Each correct answer presents a complete solution.

  NOTE: Each correct selection is worth one point

  A. Azure Blob storage container through a registered datastore

  B. Azure SQL Database

  C. URL via HTTP

  D. Azure Data Lake Storage Gen2

  E. Registered dataset

  Correct Answer: BC

• Question 183:

  You create an MLflow model

  You must deploy the model to Azure Machine Learning for batch inference.

  You need to create the batch deployment.

  Which two components should you use? Each correct answer presents a complete solution.

  NOTE: Each correct selection is worth one point

  A. Compute target

  B. Kubernetes online endpoint

  C. Model files

  D. Online endpoint

  E. Environment

  Correct Answer: AC

• Question 184:

  Note: This question is part of a series of questions that present the same scenario. Each question in the series contains a unique solution that might meet the stated goals. Some question sets might have more than one correct solution, while others might not have a correct solution.

  After you answer a question in this section, you will NOT be able to return to it. As a result, these questions will not appear in the review screen.

  You plan to use a Python script to run an Azure Machine Learning experiment. The script creates a reference to the experiment run context, loads data from a file, identifies the set of unique values for the label column, and completes the

  experiment run:

  from azureml.core import Run

  import pandas as pd

  run = Run.get_context()

  data = pd.read_csv('data.csv')

  label_vals = data['label'].unique()

  # Add code to record metrics here

  run.complete()

  The experiment must record the unique labels in the data as metrics for the run that can be reviewed later.

  You must add code to the script to record the unique label values as run metrics at the point indicated by the comment.

  Solution: Replace the comment with the following code:

  run.log_table('Label Values', label_vals)

  Does the solution meet the goal?

  A. Yes

  B. No

  Correct Answer: B

  Instead use the run_log function to log the contents in label_vals:

  for label_val in label_vals: run.log('Label Values', label_val)

  Reference: https://www.element61.be/en/resource/azure-machine-learning-services-complete-toolbox-ai

• Question 185:

  You create a classification model with a dataset that contains 100 samples with Class A and 10,000 samples with Class B.

  The variation of Class B is very high.

  You need to resolve imbalances.

  Which method should you use?

  A. Partition and Sample

  B. Cluster Centroids

  C. Tomek links

  D. Synthetic Minority Oversampling Technique (SMOTE)

  Correct Answer: D

• Question 186:

  You use Azure Machine Learning to train a model based on a dataset named dataset1.

  You define a dataset monitor and create a dataset named dataset2 that contains new data.

  You need to compare dataset1 and dataset2 by using the Azure Machine Learning SDK for Python.

  Which method of the DataDriftDetector class should you use?

  A. run

  B. get

  C. backfill

  D. update

  Correct Answer: C

  A backfill run is used to see how data changes over time.

  Reference: https://docs.microsoft.com/en-us/python/api/azureml-datadrift/azureml.datadrift.datadriftdetector.datadriftdetector

• Question 187:

  You use an Azure Machine Learning workspace.

  You have a trained model that must be deployed as a web service. Users must authenticate by using Azure Active Directory.

  What should you do?

  A. Deploy the model to Azure Kubernetes Service (AKS). During deployment, set the token_auth_enabledparameter of the target configuration object to true

  B. Deploy the model to Azure Container Instances. During deployment, set the auth_enabledparameter of the target configuration object to true

  C. Deploy the model to Azure Container Instances. During deployment, set the token_auth_enabledparameter of the target configuration object to true

  D. Deploy the model to Azure Kubernetes Service (AKS). During deployment, set the auth.enabledparameter of the target configuration object to true

  Correct Answer: A

  To control token authentication, use the token_auth_enabled parameter when you create or update a deployment

  Token authentication is disabled by default when you deploy to Azure Kubernetes Service.

  Note: The model deployments created by Azure Machine Learning can be configured to use one of two authentication methods:

  key-based: A static key is used to authenticate to the web service.

  token-based: A temporary token must be obtained from the Azure Machine Learning workspace (using Azure Active Directory) and used to authenticate to the web service.

  Incorrect Answers:

  C: Token authentication isn't supported when you deploy to Azure Container Instances.

  Reference: https://docs.microsoft.com/en-us/azure/machine-learning/how-to-authenticate-web-service

• Question 188:

  After you answer a question in this section, you will NOT be able to return to it. As a result, these questions will not appear in the review screen.

  You plan to use a Python script to run an Azure Machine Learning experiment. The script creates a reference to the experiment run context, loads data from a file, identifies the set of unique values for the label column, and completes the experiment run:

  from azureml.core import Runimport pandas as pdrun = Run.get_context()data = pd.read_csv('data.csv')label_vals = data['label'].unique()# Add code to record metrics here run.complete()

  The experiment must record the unique labels in the data as metrics for the run that can be reviewed later.

  You must add code to the script to record the unique label values as run metrics at the point indicated by the comment.

  Solution: Replace the comment with the following code:

  run.log_list('Label Values', label_vals)

  Does the solution meet the goal?

  A. Yes

  B. No

  Correct Answer: A

  run.log_list log a list of values to the run with the given name using log_list.

  Example: run.log_list("accuracies", [0.6, 0.7, 0.87])

  Note:

  Data= pd.read_csv('data.csv')

  Data is read into a pandas.DataFrame, which is a two-dimensional, size-mutable, potentially heterogeneous tabular data.

  label_vals =data['label'].unique

  label_vals contains a list of unique label values.

  Reference:

  https://www.element61.be/en/resource/azure-machine-learning-services-complete-toolbox-ai

  https://docs.microsoft.com/en-us/python/api/azureml-core/azureml.core.run(class)

  https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.html

• Question 189:

  You use the Azure Machine Learning designer to create and run a training pipeline.

  The pipeline must be run every night to inference predictions from a large volume of files. The folder where the files will be stored is defined as a dataset.

  You need to publish the pipeline as a REST service that can be used for the nightly inferencing run.

  What should you do?

  A. Create a batch inference pipeline

  B. Set the compute target for the pipeline to an inference cluster

  C. Create a real-time inference pipeline

  D. Clone the pipeline

  Correct Answer: A

  Azure Machine Learning Batch Inference targets large inference jobs that are not time-sensitive. Batch Inference provides cost-effective inference compute scaling, with unparalleled throughput for asynchronous applications. It is optimized for high-throughput, fire-and-forget inference over large collections of data.

  You can submit a batch inference job by pipeline_run, or through REST calls with a published pipeline.

  Reference: https://github.com/Azure/MachineLearningNotebooks/blob/master/how-to-use-azureml/machine-learning-pipelines/parallel-run/README.md

• Question 190:

  You have a Jupyter Notebook that contains Python code that is used to train a model.

  You must create a Python script for the production deployment. The solution must minimize code maintenance.

  Which two actions should you perform? Each correct answer presents part of the solution.

  NOTE: Each correct selection is worth one point.

  A. Refactor the Jupyter Notebook code into functions

  B. Save each function to a separate Python file

  C. Define a main() function in the Python script

  D. Remove all comments and functions from the Python script

  Correct Answer: AC

  C: Python main function is a starting point of any program. When the program is run, the python interpreter runs the code sequentially. Main function is executed only when it is run as a Python program.

  A: Refactoring, code style and testing

  The first step is to modularise the notebook into a reasonable folder structure, this effectively means to convert files from .ipynb format to .py format, ensure each script has a clear distinct purpose and organise these files in a coherent way.

  Once the project is nicely structured we can tidy up or refactor the code.

  Reference: https://www.guru99.com/learn-python-main-function-with-examples-understand-main.html https://towardsdatascience.com/from-jupyter-notebook-to-deployment-a-straightforward-example-1838c203a437