Browse Microsoft Certification Guides

Azure AI-900 ML Fundamentals Guide

Study Azure AI-900 ML Fundamentals: key concepts, common traps, and exam decision cues.

This chapter is where AI-900 checks whether you can keep predictive machine learning straight. Microsoft is not asking you to tune models or write training code. It is asking whether you can identify the right ML technique, recognize the role of data, and understand when Azure Machine Learning is the right platform boundary.

Current weight in the study guide

Microsoft currently weights this skill area at 15-20% of the exam.

Work this skill area in order

Lesson Focus
2.1 Regression, Classification, Clustering and Deep Learning Learn the predictive and grouping techniques AI-900 tests most often.
2.2 Features, Labels, Training and Validation Learn how data roles, validation, and overfitting show up at fundamentals depth.
2.3 Azure Machine Learning Capabilities Learn when the exam wants Azure Machine Learning instead of a prebuilt Azure AI service.

Fast routing inside this chapter

If the question is really about… Go first to…
choosing the right ML technique 2.1 Regression, Classification, Clustering and Deep Learning
features, labels, training data, validation data, or overfitting 2.2 Features, Labels, Training and Validation
experimentation, training, deployment, or model lifecycle on Azure 2.3 Azure Machine Learning Capabilities

What strong answers usually do

  • separate classification, regression, and clustering by the output type
  • recognize that deep learning and transformers are model-family ideas, not different workload categories by themselves
  • understand that validation data exists to test generalization, not to impress the training score
  • choose Azure Machine Learning when the scenario is about custom-model lifecycle work rather than ready-made AI services

In this section

Revised on Sunday, May 10, 2026
1. Workloads
3. Vision