====== Machine Learning in Azure ======

Machine Learning (ML) is a subset of artificial intelligence that enables systems to learn from data and improve over time without being explicitly programmed. Azure provides a suite of tools to develop, train, deploy, and manage ML models at scale.

===== 🧱 Key ML Concepts =====

  * **Supervised Learning** – Uses labeled data to train a model (e.g., predicting housing prices).
  * **Unsupervised Learning** – Finds patterns in unlabeled data (e.g., customer segmentation).
  * **Reinforcement Learning** – Learns through trial and error in an environment (e.g., game AI).

===== 🧰 Azure Machine Learning Components =====

**1. Azure Machine Learning Studio**  
  * A web-based platform for drag-and-drop model building.
  * Great for beginners and non-programmers.

**2. Azure ML SDK & CLI**  
  * Tools for professional developers and data scientists.
  * Allows full control over experiments, pipelines, and deployment.

**3. Datasets and Compute Targets**  
  * Easily manage datasets within Azure ML.
  * Choose between local compute, Azure virtual machines, or GPU clusters.

**4. Model Training & Evaluation**  
  * Automate model selection with **AutoML**.
  * Monitor accuracy, precision, recall, and other performance metrics.

**5. Deployment**  
  * Models can be deployed as REST APIs.
  * Supports scaling, version control, and monitoring.

===== 🛡 Responsible ML in Azure =====

  * Azure includes tools for **bias detection**, **explainability**, and **model fairness**.
  * Integration with Responsible AI dashboard for audit and transparency.

===== 🧠 In the AI-900 Exam =====

You need to:

  * Understand the ML lifecycle (data → train → evaluate → deploy).
  * Know when to use Azure ML Studio vs. Cognitive Services.
  * Identify the benefits of using Azure for end-to-end ML workflows.

===== 🔗 Related Topics =====

  * [[AI:AI900:azure_services:cognitive_services|Cognitive Services Overview]]
  * [[AI:AI900:tools:azure_ml_studio|Azure ML Studio (in detail)]]