The AI-900 exam has been retired
It was retired on June 30, 2026. It was replaced by AI-901 (Microsoft Azure AI Fundamentals). Go to the AI-901 study guide →
3.3.1. Automated Machine Learning
Automated ML (AutoML) eliminates the need to manually choose and tune algorithms. It automates the most time-consuming parts of machine learning.
What AutoML does:
- Automatically selects the best algorithm for your data
- Tunes hyperparameters (settings that affect model performance)
- Evaluates multiple models and recommends the best one
- Performs feature engineering automatically
How AutoML works:
- You provide a dataset and specify the target column
- AutoML tries multiple algorithms and configurations
- It evaluates each using cross-validation
- You receive a ranked list of models with performance metrics
- You can deploy the best model with one click
Supported task types:
| Task | What It Predicts | Example |
|---|---|---|
| Classification | Categories | Spam or not spam |
| Regression | Numbers | House price |
| Time series forecasting | Future values | Sales next month |
When to use AutoML:
- You have data but limited ML expertise
- You want to quickly prototype and compare approaches
- You need a baseline model to improve upon
- Time is limited and you need fast results
What AutoML eliminates:
- Choosing a model manually
- Training a model manually
- Evaluating a model manually
What AutoML does NOT eliminate:
- Creating an Azure resource (you still need to provision resources)
- Data preparation (you still need quality data)
- Understanding your business problem
⚠️ Exam Tip: AutoML is for people who want ML results without ML expertise. It automates algorithm selection and hyperparameter tuning, but you still need clean, prepared data.