Microsoft Azure AI Fundamentals Study Notes

AI & Machine Learning Fundamentals

Core concepts of artificial intelligence

Types of AI Workloads

Workload	Description	Azure Service
Machine Learning	Predictive models from data	Azure Machine Learning
Computer Vision	Analyze images and video	Azure AI Vision
Natural Language	Understand text and speech	Azure AI Language
Document Intelligence	Extract info from documents	Azure AI Document Intelligence
Generative AI	Create new content	Azure OpenAI Service

Machine Learning Types

Supervised Learning

Labeled data for training. Classification (categories) or Regression (numbers)

Unsupervised Learning

Find patterns in unlabeled data. Clustering groups similar items

Reinforcement Learning

Agent learns through trial and error with rewards

Key ML Concepts

Features: Input variables used for prediction (x values)
Labels: Values the model predicts (y values)
Training: Process of learning from data
Inference: Using trained model to make predictions
Validation: Testing model on data it hasn't seen

Exam Focus Areas

Classification = discrete categories (spam/not spam)
Regression = continuous numeric values (price prediction)
Clustering = grouping without labels (customer segments)
Anomaly detection = finding unusual patterns (fraud)

Azure Machine Learning

Cloud platform for ML development

Azure ML Capabilities

Feature	Description	User Type
Automated ML	Automatic model selection and tuning	Beginners
Designer	Drag-and-drop ML pipeline builder	Low-code users
Notebooks	Code-first development environment	Data scientists
MLOps	Model deployment and management	ML Engineers

ML Pipeline Stages

1. Data Preparation

Clean, transform, feature engineering

2. Model Training

Algorithm selection, hyperparameter tuning

3. Model Evaluation

Metrics: accuracy, precision, recall, F1

4. Deployment

Real-time endpoints or batch inference

Automated ML

AutoML automatically tries different algorithms and hyperparameters to find the best model. Great for beginners who don't know which algorithm to use.

Exam Focus Areas

AutoML for automatic model selection
Designer for visual drag-and-drop pipelines
Compute clusters for training, compute instances for development
Endpoints for deploying models as web services

Workload

Description

Azure Service

Machine Learning

Predictive models from data

Azure Machine Learning

Computer Vision

Analyze images and video

Azure AI Vision

Natural Language

Understand text and speech

Azure AI Language

Document Intelligence

Extract info from documents

Azure AI Document Intelligence

Generative AI

Create new content

Azure OpenAI Service

Feature

Description

User Type

Automated ML

Automatic model selection and tuning

Beginners

Designer

Drag-and-drop ML pipeline builder

Low-code users

Notebooks

Code-first development environment

Data scientists

MLOps

Model deployment and management

ML Engineers

Quick Feedback

Microsoft Azure AI Fundamentals Study Notes

Types of AI Workloads

Machine Learning Types

Supervised Learning

Unsupervised Learning

Reinforcement Learning

Key ML Concepts

Exam Focus Areas

Azure ML Capabilities

ML Pipeline Stages

1. Data Preparation

2. Model Training

3. Model Evaluation

4. Deployment

Automated ML

Exam Focus Areas

Microsoft Azure AI Fundamentals Study Notes

Types of AI Workloads

Machine Learning Types

Supervised Learning

Unsupervised Learning

Reinforcement Learning

Key ML Concepts

Exam Focus Areas

Azure ML Capabilities

ML Pipeline Stages

1. Data Preparation

2. Model Training

3. Model Evaluation

4. Deployment

Automated ML

Exam Focus Areas