DRAG DROP
-
You have an Azure Machine Learning workspace. You are running an experiment on your local computer.
You need to use MLflow Tracking to store metrics and artifacts from your local experiment runs in the workspace.
In which order should you perform the actions? To answer, move all actions from the list of actions to the answer area and arrange them in the correct order.
Show Answer
Correct Answer: Import MLflow and Workspace classes.
Load the workspace.
Retrieve the tracking URI and set the experiment name.
Start a training run and activate the MLflow logging API.
Explanation: You must first import required SDK classes, then load the Azure ML workspace. After the workspace is available, configure MLflow by setting the tracking URI and experiment. Finally, start the run so metrics and artifacts are logged to the workspace.
Question 190
You use the Azure Machine Learning Python SDK to create a batch inference pipeline.
You must publish the batch inference pipeline so that business groups in your organization can use the pipeline. Each business group must be able to specify a different location for the data that the pipeline submits to the model for scoring.
You need to publish the pipeline.
What should you do?
A. Create multiple endpoints for the published pipeline service and have each business group submit jobs to its own endpoint.
B. Define a PipelineParameter object for the pipeline and use it to specify the business group-specific input dataset for each pipeline run.
C. Define a OutputFileDatasetConfig object for the pipeline and use the object to specify the business group-specific input dataset for each pipeline run.
D. Have each business group run the pipeline on local compute and use a local file for the input data.
Show Answer
Correct Answer: B
Explanation: To allow different business groups to submit data from different locations when running the same published batch inference pipeline, the input must be configurable at runtime. A PipelineParameter enables parameterization of inputs (such as a datastore path or dataset reference) that can be supplied for each pipeline run without republishing the pipeline. Creating multiple endpoints is unnecessary, OutputFileDatasetConfig is for outputs (not inputs), and running locally does not meet the requirement of using a published pipeline.
Question 192
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 create an Azure Machine Learning pipeline named pipeline1 with two steps that contain Python scripts. Data processed by the first step is passed to the second step.
You must update the content of the downstream data source of pipeline1 and run the pipeline again.
You need to ensure the new run of pipeline1 fully processes the updated content.
Solution: Set the allow_reuse parameter of the PythonScriptStep object of both steps to False.
Does the solution meet the goal?
A. Yes
B. No
Show Answer
Correct Answer: A
Explanation: Yes. Setting allow_reuse=False on both PythonScriptStep objects forces each step to re-run regardless of cached outputs. This ensures that when the downstream data source is updated, the pipeline does not reuse previous results and fully processes the new content. While regenerate_outputs=True at submission time is an alternative, the proposed solution itself correctly meets the goal.
Question 193
HOTSPOT
-
You create an Azure Machine Learning model to include model files and a scoring script.
You must deploy the model. The deployment solution must meet the following requirements:
• Provide near real-time inferencing.
• Enable endpoint and deployment level cost estimates.
• Support logging to Azure Log Analytics.
You need to configure the deployment solution.
What should you configure? To answer, select the appropriate options in the answer area.
NOTE: Each correct selection is worth one point.
Explanation: Managed online endpoints provide near real-time inferencing, built-in endpoint and deployment cost estimates, and native integration with Azure Monitor and Log Analytics. Online deployments run from an environment packaged as a Docker image that contains the model dependencies and scoring script.
Question 194
DRAG DROP
-
You create an Azure Machine Learning workspace. You are training a classification model with no-code AutoML in Azure Machine Learning studio.
The model must predict if a client of a financial institution will subscribe to a fixed-term deposit. You must identify the feature that has the most influence on the predictions of the model for the second highest scoring algorithm. You must minimize the effort and time to identify the feature.
You need to complete the identification.
Which three actions should you perform in sequence? To answer, move the appropriate actions from the list of actions to the answer area and arrange them in the correct order.
Show Answer
Correct Answer: Select the second algorithm on the list of the automated ML job models.
Select the Explain model option.
Display the aggregate feature importance chart.
Explanation: To find the most influential feature for the second highest scoring model with minimal effort, first choose the correct ranked model, then open model explainability, and finally view the aggregate feature importance which summarizes overall feature influence.
Question 195
HOTSPOT
-
You have an Azure Machine Learning workspace.
You run the following code in a Python environment in which the configuration file for your workspace has been downloaded.
Instructions: For each of the following statements, select Yes if the statement is true. Otherwise, select No.
NOTE: Each correct selection is worth one point.
Show Answer
Correct Answer: No
Yes
No
No
Explanation: Experiment() creates the experiment if it does not exist and otherwise adds a new run.
Only observations is logged as a metric; run_time is printed, not logged.
No data.csv is created or uploaded in the code, so no output file is available.
Question 196
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 use Azure Machine Learning designer to load the following datasets into an experiment:
Dataset1 -
Dataset2 -
You need to create a dataset that has the same columns and header row as the input datasets and contains all rows from both input datasets.
Solution: Use the Add Rows module.
Does the solution meet the goal?
A. Yes
B. No
Show Answer
Correct Answer: A
Explanation: The Add Rows module in Azure Machine Learning designer is specifically designed to concatenate two datasets by appending rows. When the input datasets have the same columns and headers, Add Rows produces a single dataset that preserves the schema and includes all rows from both inputs, which meets the stated goal.
Question 197
HOTSPOT
-
You use Azure Machine Learning to implement hyperparameter tuning with a Bandit early termination policy.
The policy uses a slack_factor set to 0.1. an evaluation interval set to 1, and an evaluation delay set to 5.
You need to evaluate the outcome of the early termination policy.
What should you evaluate? To answer, select the appropriate options in the answer area.
NOTE: Each correct selection is worth one point.
Show Answer
Correct Answer: 91 percent
Every interval when metrics are reported, starting at evaluation interval 5.
Explanation: With a Bandit policy, slack_factor = 0.1 terminates runs performing worse than 1/(1+0.1) ≈ 91% of the best run. An evaluation_interval of 1 evaluates at every reported metric, and delay_evaluation = 5 means termination starts only after the 5th interval.
Question 198
You have a dataset that is stored in an Azure Machine Learning workspace.
You must perform a data analysis for differential privacy by using the SmartNoise SDK.
You need to measure the distribution of reports for repeated queries to ensure that they are balanced.
Which type of test should you perform?
A. Bias
B. Privacy
C. Accuracy
D. Utility
Show Answer
Correct Answer: A
Explanation: In the SmartNoise SDK evaluation framework, a **Bias test** specifically measures the distribution of outputs from repeated executions of the same query to ensure they are balanced and not systematically skewed. This directly matches the requirement to measure the distribution of reports for repeated queries and verify they are not unbalanced. Privacy, accuracy, and utility tests address different aspects (privacy guarantees, confidence bounds, and usefulness), not distribution balance.
Question 199
HOTSPOT
-
You are authoring a pipeline by using the Azure Machine Learning SDK for Python. You implement code to import all relevant classes, configure the workspace, and define all pipeline steps.
You need to initiate pipeline execution.
How should you complete the code? To answer, select the appropriate options in the answer area.
NOTE: Each correct selection is worth one point.
Show Answer
Correct Answer: Experiment
submit
Explanation: In Azure ML SDK v1, a pipeline is executed by submitting it to an Experiment using experiment.submit(pipeline), which returns a PipelineRun.
$19
Get all 477 questions with detailed answers and explanations
Instant download HTML + PDF delivered the moment payment clears.
Secure Stripe checkout we never see or store your card details.
7-day refund if files are defective see our refund policy.
