One of the first steps to design and implement a data science solution on Azure is creating an Azure Machine Learning workspace. Below is the outlined procedure to perform this.
I. Creation of an Azure Machine Learning Workspace
The Azure Machine Learning workspace is the core-component in the Azure Machine Learning service. It offers a centralized location where you can manage, develop, and deploy various data experiments.
Prerequisites
Before you begin, remember the following restrictions:
- You should have the necessary permissions to create a resource group and resources within your Azure subscription.
- The chosen subscription needs to have enough quota to create an Azure Machine Learning workspace and dependent resources.
- In your Azure environment, your Azure Active Directory tenant needs to be linked to your subscription.
Steps
To create an Azure Machine Learning workspace, perform the following steps:
- Sign in to the Azure portal.
- Click on the “Create a resource” option, in the menu on the left.
- Search for Machine Learning.
- Select Machine Learning.
- On the Machine Learning pane, which will appear on your screen, select the subscription where you want to create the Machine Learning workspace.
- For the Resource group, make a selection on whether to create a new one or use an existing one.
- Fill in the workspace name, region, and storage account details.
- After confirming all details are correct, click create.
II. Verifying the Workspace Installation
Once the above process is complete, you can check your workspace. Navigate to the resource group where you chose to deploy the workspace. If the deployment was successful, your workspace will be listed as one of the resources under the resource group.
III. Accessing the Workspace
To access the workspace, click on the listed workspace resource. You will receive an overview of the workspace, where you can view the associated resources such as experiments, pipelines, models, endpoints, and compute resources.
IV. Developing and Running an Experiment
After creating the workspace, you can now develop and run an experiment. Azure Machine Learning studio provides a central place for this data science workflow. You can manage, develop, and track machine learning applications from early ideation to deployment within the studio.
Conclusion
In conclusion, creating an Azure Machine Learning workspace is the first step you need to take to run machine learning tasks in the Azure environment. Understanding how to create and manage the workspace, including its configuration and setup process, is vital. Knowledge of Azure Machine Learning workspace creation/setup is also necessary for taking the “DP-100 Designing and Implementing a Data Science Solution on Azure” certification exam.
Practice Test
True/False: Azure Machine Learning workspace is the top-level resource for Azure Machine Learning, providing a centralized place to work with all the artifacts you create.
- True
- False
Answer: True
Explanation: Azure Machine Learning workspace is indeed the top-level resource for Azure Machine Learning, providing a centralized place to work with all the artifacts you create.
In creating an Azure Machine Learning workspace, which of the following is not required?
- a) Subscription
- b) Resource Group
- c) Azure Storage Account
- d) Workspace name
Answer: c) Azure Storage Account
Explanation: Azure storage account is not directly required in creating Azure ML workspace. You only need a subscription, a resource group, and a workspace name to create one.
True/False: The workspace name in Azure Machine Learning has global uniqueness across Azure.
- True
- False
Answer: False
Explanation: The workspace name does not need global uniqueness, only within the resource group within Azure.
Multiple select: Which of the following features are provided by Azure Machine Learning workspace?
- a) A central place to manage resources
- b) Converting data into useful formats
- c) Building, deploying, and managing models
- d) Handling network traffic
Answer: a) A central place to manage resources, c) Building, deploying, and managing models
Explanation: Azure Machine Learning workspace provides features like managing resources and building, deploying, and managing models. It does not handle network traffic or convert data into useful formats directly.
True/False: You can create multiple workspaces in the same resource group in Azure.
- True
- False
Answer: True
Explanation: Azure allows you to create multiple workspaces within the same resource group.
What is the maximum number of characters allowed in an Azure ML workspace name?
- a) 64
- b) 32
- c) 16
- d) 128
Answer: b) 32
Explanation: An Azure ML workspace name can have up to 32 characters.
True/False: Azure ML workspace needs to be linked with an Azure Databricks workspace at the time of creation.
- True
- False
Answer: False
Explanation: It’s not mandatory to link Azure ML workspace with an Azure Databricks workspace at the time of creation. You can link them anytime later.
After creating an Azure ML workspace, where do you go to monitor costs?
- a) Azure portal
- b) Azure ML Studio
- c) Azure DevOps
- d) Physically go to the Azure data center
Answer: a) Azure portal
Explanation: You can monitor the costs of your Azure ML workspace from the Azure portal.
Can you rename the Azure Machine Learning workspace after it’s created?
- a) Yes
- b) No
Answer: b) No
Explanation: Once you’ve created an Azure Machine Learning workspace, you can’t rename it.
True/False: The Azure Machine Learning workspace allows you to manage resources and collaborations in one central location.
- True
- False
Answer: True
Explanation: Yes, Azure Machine Learning workspace does enable you to manage resources and collaborations in a centralized location.
Single select: Who can create an Azure Machine Learning workspace?
- a) Any active Azure user
- b) Only users with a Machine Learning role
- c) Only a DevOps engineer
- d) Only the Azure subscription owner
Answer: a) Any active Azure user
Explanation: Any active Azure user can create an Azure Machine Learning workspace, not limited to specific roles only.
Is linking the Azure Machine Learning workspace to an Azure Container Registry mandatory?
- a) Yes
- b) No
Answer: b) No
Explanation: It is not mandatory to link the Azure Machine Learning workspace to an Azure Container Registry – this is optional.
True/False: Azure Machine Learning workspace creation is irreversible.
- True
- False
Answer: False
Explanation: You can delete the Azure Machine Learning workspace you created if it’s not needed anymore.
What happens to associated resources when you delete an Azure ML Workspace?
- a) They get deleted too
- b) They remain intact
- c) It depends on the type of resource
- d) They get transferred to another workspace
Answer: b) They remain intact
Explanation: When you delete a workspace, only the workspace is deleted. The associated resources such as storage accounts, compute instances, etc. remain intact.
True/False: Azure Machine Learning workspace supports both private and public endpoint types.
- True
- False
Answer: True
Explanation: Yes, Azure Machine Learning workspace supports both private endpoint (Azure Private Link) and public endpoint types.
Interview Questions
What is the first requirement to create an Azure Machine Learning workspace?
The initial requirement is that you need to have an Azure subscription.
Which components are created when an Azure Machine Learning workspace is made?
When creating an Azure Machine Learning workspace, an Azure Storage account, an Azure Key Vault, an Azure Container Registry, and an Azure Application Insights resource is also created.
Once an Azure Machine Learning workspace is created can you change its associated resources like the Key Vault, Storage, Application Insights, etc.?
No, once a workspace is created, you cannot change its associated resources.
Is the Azure Machine Learning Studio free to use?
No, costs are associated with running computations and storing data in Azure Machine Learning.
Why would you need to create more than one Azure Machine Learning workspace?
Creating multiple Machine Learning workspaces can be beneficial for separating different project environments or separating environments such as production, development, and testing.
In which five locations can we establish an Azure Machine Learning workspace?
Azure Machine Learning workspace can be established in the Azure portal, Azure Machine Learning studio, Azure Resource Manager template, Azure CLI, and Azure PowerShell.
Can you access data in Azure Synapse Analytics from your Azure Machine Learning workspace?
Yes, you can access data in Azure Synapse Analytics from Azure Machine Learning studio.
What is the importance of an Azure Machine Learning workspace?
An Azure Machine Learning workspace is a foundational resource in the cloud that you use to experiment, train, and deploy machine learning models.
What happens if you delete a Machine learning workspace?
If you delete a Machine learning workspace, you also delete all the resources associated with it, e.g., models, experiments, deployments, etc.
Is it possible to add collaborators to your Azure Machine Learning workspace?
Yes, you can add collaborators by assigning them a specific role at the workspace level.
Can we create a private endpoint for an Azure Machine Learning workspace?
Yes, when a workspace is enabled with a private endpoint, you have a secure and direct connection over a private network.
What is the relationship between Azure machine learning workspace and resource group?
An Azure machine learning workspace is a resource that resides in an Azure resource group, which is a container that holds related resources for an Azure solution.
Can I move my Azure Machine Learning workspace to another subscription or resource group?
Yes, you can move your workspace to a different subscription or resource group.
Can I enforce network isolation for my Azure Machine Learning workspace?
Yes, you can enable Azure Private Link to divert the network traffic between your workspace and Azure Machine Learning control plane over a private network connection.
Which tools can I use to interact with the Azure Machine Learning workspace?
You can interact with the Azure Machine Learning workspace via the Azure Machine Learning SDKs, REST APIs, Azure Machine Learning studio interface or other Azure services that integrate with Azure Machine Learning.
extutorials.com