HOTSPOT
-
You use Azure Machine Learning to train a machine learning model.
You use the following training script in Python to perform logging:
import mlflow
mlflow.log_metric(“accuracy", float(vel_accuracy))
You must use a Python script to define a sweep job.
You need to provide the primary metric and goal you want hyperparameter tuning to optimize.
How should you complete the Python script? To answer, select the appropriate options in the answer area.
NOTE: Each correct selection is worth one point.
Explanation: The training script logs a metric named "accuracy" using mlflow.log_metric. Hyperparameter tuning should therefore optimize this metric, and since higher accuracy is better, the goal is to maximize it.
Question 119
DRAG DROP
-
You manage an Azure Machine Learning workspace. You train a model named model1.
You must identify the features to modify for a differing model prediction result.
You need to configure the Responsible AI (RAI) dashboard for model1.
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: Load and configure the Responsible AI Insights dashboard constructor component.
Add the counterfactuals component to the Responsible AI Insights dashboard.
Use the Gather Responsible AI Insights dashboard component to present the dashboard.
Explanation: To identify which features must change to obtain a different prediction, the Counterfactuals component is required. The RAI Insights dashboard constructor initializes the dashboard, at least one analysis tool (counterfactuals) is added, and the Gather component publishes the final dashboard.
Question 120
You have a dataset that contains records of patients tested for diabetes. The dataset includes the patient's age.
You plan to create an analysis that will report the mean age value from the differentially private data derived from the dataset.
You need to identify the epsilon value to use in the analysis that minimizes the risk of exposing the actual data.
Which epsilon value should you use?
A. -1.5
B. -0.5
C. 0.5
D. 1.5
Show Answer
Correct Answer: C
Explanation: In differential privacy, the privacy budget ε must be non-negative, and smaller values provide stronger privacy guarantees by adding more noise. Negative ε values are invalid. Between the valid options (0.5 and 1.5), ε = 0.5 minimizes the risk of exposing the actual data.
Question 121
HOTSPOT
-
You monitor an Azure Machine Learning classification training experiment named train.classification on Azure Notebooks.
You must store a table named table as an artifact in Azure Machine Learning Studio during model training.
You need to collect and list the metrics by using MLflow.
How should you complete the code segment? To answer, select the appropriate option in the answer area.
NOTE: Each correct selection is worth one point.
Explanation: Metrics are logged as key–value pairs using mlflow.log_metrics. Files such as tables must be saved locally and logged with mlflow.log_artifact. To retrieve the completed run, MlflowClient.get_run requires the current run ID, available as mlflow_run.info.run_id.
Question 123
You create an Azure Machine learning workspace.
You must use the Azure Machine Learning Python SDK v2 to define the search space for discrete hyperparameters. The hyperparameters must consist of a list of predetermined, comma-separated integer values.
You need to import the class from the azure.ai.ml.sweep package used to create the list of values.
Which class should you import?
A. Choice
B. Randint
C. Uniform
D. Normal
Show Answer
Correct Answer: A
Explanation: In Azure Machine Learning SDK v2, discrete hyperparameters defined as a fixed list of predetermined values use the Choice class from azure.ai.ml.sweep. Randint, Uniform, and Normal are for sampled numeric distributions rather than explicit lists of values.
Question 124
You use Azure Machine Learning to train a model.
You must use Bayesian sampling to tune hyperparameters.
You need to select a learning_rate parameter distribution.
Which two distributions can you use? Each correct answer presents a complete solution.
NOTE: Each correct selection is worth one point.
A. Uniform
B. Choice
C. QNormal
D. Normal
E. LogUniform
Show Answer
Correct Answer: A, B
Explanation: In Azure Machine Learning, Bayesian sampling supports only a limited set of parameter distributions: choice, uniform, and quniform. For a learning_rate parameter under Bayesian sampling, Uniform and Choice are valid options. Normal, QNormal (different from quniform), and LogUniform are not supported with Bayesian sampling.
Question 125
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 train and register an Azure Machine Learning model.
You plan to deploy the model to an online endpoint.
You need to ensure that applications will be able to use the authentication method with a non-expiring artifact to access the model.
Solution: Create a managed online endpoint with the default authentication settings. Deploy the model to the online endpoint.
Does the solution meet the goal?
A. Yes
B. No
Show Answer
Correct Answer: A
Explanation: A managed online endpoint uses key-based authentication by default. Endpoint keys are non-expiring unless manually regenerated, which satisfies the requirement for applications to authenticate using a non-expiring artifact. Therefore, creating a managed online endpoint with default authentication settings meets the goal.
Question 126
HOTSPOT -
You are using hyperparameter tuning in Azure Machine Learning Python SDK v2 to train a model.
You configure the hyperparameter tuning experiment by running the following code:
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: The sweep uses random/continuous sampling, not an exhaustive grid, so not all combinations are tested.
learning_rate=Normal(10, 3) samples from a normal distribution with mean 10 and standard deviation 3.
keep_probability=Uniform(0.05, 0.1) samples any value in the range, not just the endpoints.
number_of_hidden_layers=Choice(range(3,5)) selects discrete values (3 or 4), not a normal distribution.
Question 127
DRAG DROP
-
You manage an Azure Machine Learning workspace named workspace1 and a Data Science Virtual Machine (DSVM) named DSMV1.
You must run an experiment on DSMV1 by using a Jupyter notebook and Python SDK v2 code. You must store metrics and artifacts in workspace1. You start by creating Python SDK v2 code to import all required packages.
You need to implement the Python SDK v2 code to store metrics and artifacts in workspace1.
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: Instantiate an object of the MLClient class.
Retrieve the tracking URI of workspace1.
Set the MLflow tracking URI.
Explanation: In Azure ML SDK v2, MLClient is required to connect to the workspace. The workspace’s MLflow tracking URI is then retrieved through the client and configured in MLflow so that all metrics and artifacts logged from the Jupyter notebook are stored in the Azure Machine Learning workspace.
Question 128
HOTSPOT
-
You use Azure Machine Learning and SmartNoise Python libraries to implement a differential privacy solution to protect a dataset containing citizen demographics for the city of Seattle in the United States.
The solution has the following requirements:
• Allow for multiple queries targeting the mean and variance of the citizen’s age.
• Ensure full plausible deniability.
You need to define the query rate limit to minimize the risk of re-identification.
What should you configure? To answer, select the appropriate options in the answer area.
NOTE: Each correct selection is worth one point.
Show Answer
Correct Answer: Privacy budget
Set the epsilon value to the smaller of the epsilon values assigned to the mean and variance queries
Explanation: In differential privacy, query rate limiting is enforced through the privacy budget (ε). To minimize re-identification risk and ensure strong plausible deniability, the system should cap queries using the most restrictive (smallest) epsilon, which limits cumulative privacy loss across multiple statistical queries.
$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.
