Microsoft Microsoft Certifications DP-100 Questions & Answers

• Question 371:

  You plan to build a team data science environment. Data for training models in machine learning pipelines will be over 20 GB in size. You have the following requirements:

  1.

  Models must be built using Caffe2 or Chainer frameworks.

  2.

  Data scientists must be able to use a data science environment to build the machine learning pipelines and train models on their personal devices in both connected and disconnected network environments.

  Personal devices must support updating machine learning pipelines when connected to a network.

  You need to select a data science environment.

  Which environment should you use?

  A. Azure Machine Learning Service

  B. Azure Machine Learning Studio

  C. Azure Databricks

  D. Azure Kubernetes Service (AKS)

  Correct Answer: A

  The Data Science Virtual Machine (DSVM) is a customized VM image on Microsoft's Azure cloud built specifically for doing data science. Caffe2 and Chainer are supported by DSVM. DSVM integrates with Azure Machine Learning.

  Incorrect Answers:

  B: Use Machine Learning Studio when you want to experiment with machine learning models quickly and easily, and the built-in machine learning algorithms are sufficient for your solutions.

  References: https://docs.microsoft.com/en-us/azure/machine-learning/data-science-virtual-machine/overview

• Question 372:

  You manage an Azure Machine Learning workspace. The workspace includes an Azure Machine Learning Kubernetes compute target configured as an Azure Kubernetes Service (AKS) cluster named AKS1. AKS1 is configured to enable the

  targeting of different nodes to train workloads.

  You must run a command job on AKS1 by using the Azure ML Python SDK v2. The command job must select different types of compute nodes. The compute node types must be specified by using a command parameter.

  You need to configure the command parameter.

  Which parameter should you use?

  A. environment

  B. compute

  C. limits

  D. instance_type

  Correct Answer: D

  command

  Create a Command object which can be used inside dsl.pipeline as a function and can also be created as a standalone command job.

  Parameters include:

  instance_type

  Optional type of VM used as supported by the compute target.

  Incorrect:

  *

  environment

  the environment to use for this command

  *

  compute

  the name of the compute where the command job is executed( will not be used if the command is used as a component/function)

  Reference:

  https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.m

• Question 373:

  You manage an Azure Machine Learning workspace named workspace1.

  You must develop Python SDK v2 code to add a compute instance to workspace1. The code must import all required modules and call the constructor of the ComputeInstance class.

  You need to add the instantiated compute instance to workspace1.

  What should you use?

  A. constructor of the azure.ai.ml.ComputeSchedule class

  B. constructor of the azure.ai.ml.ComputePowerAction enum

  C. begin_create_or_update method of an instance of the azure.ai.ml.MLCIient class

  D. set_resources method of an instance of the azure.ai.ml.Command class

  Correct Answer: C

  Creating a compute instance is a one time process for your workspace. You can reuse the compute as a development workstation or as a compute target for training. You can have multiple compute instances attached to your workspace.

  Example: # Compute Instances need to have a unique name across the region. # Here we create a unique name with current datetime from azure.ai.ml.entities import ComputeInstance, AmlCompute import datetime

  ci_basic_name = "basic-ci" + datetime.datetime.now().strftime("%Y%m%d%H%M") ci_basic = ComputeInstance(name=ci_basic_name, size="STANDARD_DS3_v2") ml_client.begin_create_or_update(ci_basic).result()

  For more information on the classes, methods, and parameters used in this example, see the following reference documents:

  AmlCompute class ComputeInstance class

  Reference: https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.ml.entities.computeinstance

• Question 374:

  You create a workspace to include a compute instance by using Azure Machine Learning Studio. You are developing a Python SDK v2 notebook in the workspace.

  You need to use Intellisense in the notebook.

  What should you do?

  A. Stop the compute instance.

  B. Start the compute instance.

  C. Run a %pip magic function on the compute instance.

  D. Run a !pip magic function on the compute instance.

  Correct Answer: C

  Manage packages

  Since your compute instance has multiple kernels, make sure use %pip or %conda magic functions, which install packages into the currently running kernel. Don't use !pip or !conda, which refers to all packages (including packages outside

  the currently running kernel).

  Note: Code completion (IntelliSense)

  IntelliSense is a code-completion aid that includes many features: List Members, Parameter Info, Quick Info, and Complete Word. With only a few keystrokes, you can:

  Learn more about the code you're using

  Keep track of the parameters you're typing

  Add calls to properties and methods

  Reference:

  https://learn.microsoft.com/en-us/azure/machine-learning/how-to-run-jupyter-notebooks

  https://code.visualstudio.com/docs/editor/intellisense

• Question 375:

  You manage an Azure Machine Learning workspace.

  You need to define an environment from a Docker image by using the Azure Machine Learning Python SDK v2.

  Which parameter should you use?

  A. properties

  B. image

  C. build

  D. conda_file

  Correct Answer: B

  Create an environment from a Docker image

  To define an environment from a Docker image, provide the image URI of the image hosted in a registry such as Docker Hub or Azure Container Registry.

  The following example is a YAML specification file for an environment defined from a Docker image. An image from the official PyTorch repository on Docker Hub is specified via the image property in the YAML file.

  $schema: https://azuremlschemas.azureedge.net/latest/environment.schema.json

  name: docker-image-example

  image: pytorch/pytorch:latest

  description: Environment created from a Docker image.

  To create the environment:

  az ml environment create --file assets/environment/docker-image.yml

  Reference:

  https://learn.microsoft.com/en-us/azure/machine-learning/how-to-manage-environments-v2

• Question 376:

  You create an Azure Machine Learning managed compute resource. The compute resource is configured as follows:

  Minimum nodes: 2 Maximum nodes: 4 You must decrease the minimum number of nodes and increase the maximum number of nodes to the following values: Minimum nodes: 0

  Maximum nodes: 8

  You need to reconfigure the compute resource.

  Which three methods can you use? Each correct answer presents a complete solution.

  NOTE: Each correct selection is worth one point.

  A. Azure Machine Learning designer

  B. MLClient class in Python SDK v2

  C. Azure Machine Learning studio

  D. Azure CLI ml extension v2

  E. BuildContext class in Python SDK v2

  Correct Answer: ACD

  Incorrect:

  *

  MLClient Class

  A client class to interact with Azure ML services.

  Use this client to manage Azure ML resources, e.g. workspaces, jobs, models and so on.

  Constructor

  MLClient(credential: TokenCredential, subscription_id: str | None = None, resource_group_name: str | None = None, workspace_name: str | None = None, registry_name: str | None = None, **kwargs: Any)

  *

  DockerBuildContext Class

  Defines a Docker build context.

  Class DockerBuildContext constructor.

  Constructor

  DockerBuildContext(location_type=None, location=None, dockerfile_path='Dockerfile')

  Reference: https://github.com/MicrosoftDocs/azure-docs/blob/master/articles/machine-learning/how-to-setup-authentication.md https://learn.microsoft.com/en-us/python/api/azure-ai-ml/azure.ai.ml.entities.buildcontext

• Question 377:

  You plan to use automated machine learning by using Azure Machine Learning Python SDK v2 to train a regression model. You have data that has features with missing values, and categorical features with few distinct values.

  You need to control whether automated machine learning automatically imputes missing values and encode categorical features as part of the training task.

  Which enum of the automl package should you use?

  A. ForecastHorizonMode

  B. RegressionModels

  C. FeaturizationMode

  D. RegressionPrimaryMetrics

  Correct Answer: C

  AutoMLConfig Class

  Represents configuration for submitting an automated ML experiment in Azure Machine Learning.

  This configuration object contains and persists the parameters for configuring the experiment run, as well as the training data to be used at run time.

  Constructor

  Featurization str or FeaturizationConfig

  'auto' / 'off' / FeaturizationConfig Indicator for whether featurization step should be done automatically or not, or whether customized featurization should be used.

  Column type is automatically detected. Based on the detected column type preprocessing/featurization is done as follows:

  Categorical: Target encoding, one hot encoding, drop high cardinality categories, impute missing values.

  *-> Numeric: Impute missing values, cluster distance, weight of evidence.

  DateTime: Several features such as day, seconds, minutes, hours etc.

  Text: Bag of words, pre-trained Word embedding, text target encoding.

  Reference:

  https://learn.microsoft.com/en-us/python/api/azureml-train-automl-client/azureml.train.automl.automlconfig.automlconfig https://docs.microsoft.com/en-us/azure/machine-learning/how-to-use-automated-ml-for-ml-models

• Question 378:

  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 Join Data module.

  Does the solution meet the goal?

  A. Yes

  B. No

  Correct Answer: B

  Correct Solution: Use the Add Rows module.

  Note: Add Rows component

  Use this component to concatenate two datasets. In concatenation, the rows of the second dataset are added to the end of the first dataset.

  Concatenation of rows is useful in scenarios such as these:

  You have generated a series of evaluation statistics, and you want to combine them into one table for easier reporting.

  You have been working with different datasets, and you want to combine the datasets to create a final dataset.

  Reference: https://learn.microsoft.com/en-us/azure/machine-learning/component-reference/add-rows

• Question 379:

  You are developing a two-step Azure Machine Learning pipeline by using the Azure Machine Learning SDK for Python. You need to register the output of the pipeline as a new version of a named dataset after the run has been completed. What should you implement?

  A. the as_input method of the OutputDatasetConfig class

  B. the register_on_complete method of the OutputDatasetConfig class

  C. the as_mount method of the DatasetConsumptionConfig class

  D. the as_download method of the DatasetConsumptionConfig class

  Correct Answer: A

  The OutputDatasetConfig Class as_input method registers the output as a new version of a named Dataset after the run has ran.

  If there are no datasets registered under the specified name, a new Dataset with the specified name will be registered. If there is a dataset registered under the specified name, then a new version will be added to this dataset.

  Incorrect:

  *

  The OutputDatasetConfig Class as_input method specifies how to consume the output as an input in subsequent pipeline steps.

  *

  as_mount sets the mode to mount.

  In the submitted run, files in the datasets will be mounted to local path on the compute target. The mount point can be retrieved from argument values and the input_datasets field of the run context.

  *

  as_download Set the mode to download.

  In the submitted run, files in the dataset will be downloaded to local path on the compute target. The download location can be retrieved from argument values and the input_datasets field of the run context.

  Reference: https://learn.microsoft.com/en-us/python/api/azureml-core/azureml.data.output_dataset_config.outputdatasetconfig https://learn.microsoft.com/en-us/python/api/azureml-core/azureml.data.dataset_consumption_config.datasetconsumptionconfig

• Question 380:

  You are implementing hyperparameter tuning by using Bayesian sampling for a model training from a notebook. The notebook is in an Azure Machine Learning workspace that uses a compute cluster with 20 nodes.

  The code implements Bandit termination policy with slack factor set to 0.2 and the HyperDriveConfig class instance with max_concurrent_runs set to 10.

  You must increase effectiveness of the tuning process by improving sampling convergence.

  You need to select which sampling convergence to use.

  What should you select?

  A. Set the value of slack factor of early_termination_policy to 0.9.

  B. Set the value of max_concurrent_runs of HyperDriveConfig to 4.

  C. Set the value of slack factor of early_termination_policy to 0.1.

  D. Set the value of max_concurrent_runs of HyperDriveConfig to 20.

  Correct Answer: B

  Bayesian sampling

  Bayesian sampling is based on the Bayesian optimization algorithm. It picks samples based on how previous samples did, so that new samples improve the primary metric.

  The number of concurrent jobs has an impact on the effectiveness of the tuning process. A smaller number of concurrent jobs may lead to better sampling convergence, since the smaller degree of parallelism increases the number of jobs

  that benefit from previously completed jobs.

  Reference: https://learn.microsoft.com/en-us/azure/machine-learning/how-to-tune-hyperparameters