Azure AI Machine Learning - Overview
Welcome to the comprehensive reference documentation for Azure AI Machine Learning. This platform provides a unified cloud service for building, training, deploying, and managing machine learning models at scale.
Key Capabilities
Azure AI Machine Learning empowers data scientists and developers with tools and services to accelerate their ML workflows, from experimentation to production.
Core Components and Concepts
Azure AI Machine Learning is built around several key concepts that enable efficient and scalable machine learning development:
Workspaces
An Azure AI Machine Learning workspace is the top-level resource for Azure AI Machine Learning activities. It provides a centralized place to manage all artifacts that you create, such as datasets, models, experiments, and compute resources.
Compute Resources
You can create and manage various compute resources within your workspace, including:
- Compute Instances: Cloud-based workstations for development and testing.
- Compute Clusters: Scalable clusters for training and batch inference.
- Inference Clusters: Kubernetes clusters for deploying models as real-time web services.
Datasets and Datastores
Easily manage and version your data. Datastores securely connect to your storage services (like Azure Blob Storage, Azure Data Lake Storage), and Datasets provide an abstraction to easily access and track your data.
Models
Register, version, and manage your trained machine learning models. Azure AI Machine Learning supports various model formats and provides tools for tracking model lineage and metadata.
Environments
Define and manage reproducible Python environments for your training and inference jobs. This ensures consistency across development and production.
Jobs and Experiments
Submit and track your training runs as Jobs, which are grouped into Experiments. Monitor metrics, visualize results, and compare different runs to find the best performing models.
Endpoints
Deploy your trained models as web services using managed online endpoints for real-time inference or batch endpoints for batch scoring.
Key Features at a Glance
Data Management
Version control, access, and manage datasets with ease.
Model Training
Build, train, and tune models using various ML frameworks.
Model Deployment
Deploy models as scalable web services for real-time or batch inference.
MLOps
Implement MLOps practices with integrated CI/CD, monitoring, and governance.
Responsible AI
Tools for fairness, explainability, and privacy-preserving ML.
Scalable Infrastructure
Leverage Azure's robust and scalable cloud infrastructure.
Getting Started
To begin your journey with Azure AI Machine Learning, explore the following resources:
- Azure AI Machine Learning SDK to interact with the service programmatically.
- Tutorials for guided walkthroughs of common ML tasks.
- Dataset management guides to learn how to prepare and use your data.
This documentation serves as your central hub for all things Azure AI Machine Learning. Navigate through the sections to find detailed guides, API references, and best practices.