Getting Started with Azure AI Machine Learning

Welcome to the Azure AI Machine Learning documentation. This guide will walk you through the essential steps to begin your journey with our powerful cloud-based platform for building, training, and deploying machine learning models at scale.

What is Azure AI Machine Learning?

Azure AI Machine Learning is a fully managed cloud service that you can use to train, deploy, manage, and monitor machine learning models. It provides a comprehensive environment for data scientists and developers to build and manage their ML lifecycle. Key features include:

Data preparation and transformation
Model training and hyperparameter tuning
Model deployment to various endpoints
MLOps for lifecycle management
Responsible AI tools

Prerequisites

Before you begin, ensure you have the following:

An Azure account. If you don't have one, you can sign up for a free trial.
An Azure AI Machine Learning workspace.

Step 1: Create an Azure AI Machine Learning Workspace

Sign in to the Azure portal

Navigate to the Azure portal and sign in with your Azure account credentials.

Create a new Azure AI Machine Learning resource

In the Azure portal, search for "Azure AI Machine Learning" and select "Create". Follow the prompts to configure your workspace, including resource group, region, and workspace name.

Configure storage and networking (optional)

You can choose to use default settings or configure custom storage accounts and virtual networks for enhanced security and control.

Tip: Workspaces are the foundational resource for Azure AI Machine Learning. They organize all your assets and provide access control.

Step 2: Set up Your Development Environment

You have several options for interacting with your Azure AI Machine Learning workspace:

Option 1: Azure Machine Learning Studio (Web Interface)

The Azure Machine Learning studio is a web-based portal that provides a user-friendly interface for managing your ML projects. It includes tools for data preparation, model training (Designer), automated ML, and notebook integration.

To access it, go to ml.azure.com and sign in with your Azure account.

Option 2: Azure Machine Learning SDK (Python)

For programmatic access and automation, use the Azure Machine Learning SDK for Python. This allows you to interact with your workspace using Python scripts.

Installation:

pip install azure-ai-ml azure-identity

Authentication and Workspace Connection:


from azure.ai.ml import MLClient
from azure.identity import DefaultAzureCredential

# Authenticate using Azure Identity
credential = DefaultAzureCredential()

# Replace with your subscription ID, resource group, and workspace name
subscription_id = "YOUR_SUBSCRIPTION_ID"
resource_group = "YOUR_RESOURCE_GROUP"
workspace_name = "YOUR_WORKSPACE_NAME"

ml_client = MLClient(
    credential=credential,
    subscription_id=subscription_id,
    resource_group=resource_group,
    workspace_name=workspace_name,
)

print(f"Connected to workspace: {ml_client.workspace_name}")

Option 3: Azure CLI Extension

The Azure CLI provides commands to manage your Azure AI Machine Learning resources from the command line.

Installation:

az extension add -n ml

Login and Set Default Workspace:


az login
az account set --subscription "YOUR_SUBSCRIPTION_ID"
az configure --defaults group="YOUR_RESOURCE_GROUP" workspace="YOUR_WORKSPACE_NAME"

Step 3: Your First ML Project

Once your workspace is set up and your development environment is configured, you can start your first machine learning project. This typically involves:

Uploading or connecting to your data.
Choosing a model training method (e.g., Automated ML, Designer, custom scripts).
Training your model.
Evaluating and registering your model.
Deploying your model as a web service.

Next Steps: Explore the Tutorials section to follow step-by-step guides for common ML scenarios.

We are excited to have you on board with Azure AI Machine Learning. Happy modeling!