Azure AI & Machine Learning Documentation

Welcome to the comprehensive documentation for Azure AI and Machine Learning. This guide will help you understand and utilize the vast array of services and tools available to build intelligent applications.

Comprehensive Capabilities

From pre-built AI models to custom ML development, Azure covers your entire AI lifecycle.

Scalable Infrastructure

Leverage Azure's robust cloud infrastructure for training and deploying models at any scale.

Responsible AI

Build trustworthy AI systems with built-in tools for fairness, transparency, and privacy.

Integrated Services

Seamlessly integrate AI and ML capabilities into your existing applications and workflows.

Getting Started

Begin your journey with Azure AI and Machine Learning by setting up your environment and exploring the fundamental concepts.

Key Steps:

  • Create an Azure Subscription and a resource group.
  • Provision an Azure Machine Learning Workspace, the central hub for your ML projects.
  • Familiarize yourself with the Azure ML Studio, a web-based IDE for ML tasks.
  • Understand core concepts like Datasets, Compute Targets, and Experiments.

For a hands-on introduction, check out our Getting Started Tutorial.

Azure Machine Learning Core Services

Azure Machine Learning provides a set of interconnected services to manage the end-to-end machine learning lifecycle.

Azure ML Workspace

The Azure ML Workspace is the foundational resource for all your machine learning activities. It provides a centralized location to manage experiments, models, datasets, and compute resources.

Datasets and Datastores

Organize and access your data efficiently. Datastores securely connect to your data in Azure storage (Blob Storage, Azure Data Lake Storage, etc.), while Datasets provide a versioned view of your data for training and evaluation.

# Example of registering a dataset
from azureml.core import Workspace, Dataset

ws = Workspace.from_config()
dataset = Dataset.Tabular.from_delimited_files(path='path/to/your/data.csv')
dataset = dataset.register(workspace=ws,
                           name='my-training-data',
                           description='Training data for model X')

Compute Targets

Select the right compute infrastructure for your needs. Azure ML supports various compute targets, including CPU and GPU virtual machines, managed compute clusters, and Kubernetes.

Models

Register, version, and manage your trained machine learning models. Once registered, models can be easily deployed to various endpoints.

Endpoints

Deploy your trained models as scalable web services. Azure ML supports both real-time inference endpoints (for low-latency predictions) and batch inference endpoints (for processing large datasets).

Azure AI Services

Beyond custom ML, Azure offers a rich portfolio of pre-built AI services that can be easily integrated into your applications.

Cognitive Services

Empower your applications with human-like intelligence through Cognitive Services. This includes Vision, Speech, Language, Decision, and Search capabilities.

  • Vision: Image analysis, object detection, facial recognition.
  • Speech: Speech to text, text to speech, speaker recognition.
  • Language: Sentiment analysis, key phrase extraction, language understanding.
  • Decision: Anomaly detector, content moderator, personalizer.

Azure OpenAI

Access powerful large language models like GPT-3.5 and GPT-4, fine-tuned for enterprise use with enhanced security, privacy, and responsible AI features.

Explore capabilities for content generation, summarization, code generation, and more.

Bot Service

Build, connect, test, and deploy intelligent bots that interact with users naturally across various channels.

Tutorials and Quickstarts

Dive into practical examples and step-by-step guides to get hands-on experience with Azure AI and Machine Learning.

Best Practices

Learn how to build robust, scalable, and maintainable AI solutions on Azure.

  • Effective data management and versioning.
  • Choosing appropriate compute resources.
  • Model monitoring and retraining strategies.
  • Implementing Responsible AI principles.
  • Securing your AI workloads.

API Reference

Access detailed documentation for all Azure AI and Machine Learning SDKs and REST APIs.