Artificial Intelligence (AI), Machine Learning (ML), AI, Microsoft Certification, AI-900

Identify Regression Machine Learning Scenarios (AI-900 Exam Prep)

Where This Fits in the Exam

  • Exam Domain: Describe fundamental principles of machine learning on Azure (15–20%)
  • Sub-Domain: Identify common machine learning techniques
  • Topic: Identify regression machine learning scenarios

On the AI-900 exam, regression questions are about recognizing when regression is the appropriate technique, not building or tuning models.

What Is Regression in Machine Learning?

Regression is a type of supervised machine learning used to predict a numerical (continuous) value.

  • The model learns from labeled training data
  • The output is a number, not a category
  • The goal is to predict how much, how many, or how long

Key exam rule:
If the output is a number, the scenario is almost always regression.

Characteristics of Regression Scenarios

A regression machine learning workload typically involves:

  • Historical data with known outcomes
  • One or more input features
  • A continuous numeric output
  • Predicting future values based on patterns in data

Examples of numeric outputs:

  • Price
  • Temperature
  • Revenue
  • Distance
  • Duration
  • Quantity

Common Regression Use Cases

Price and Cost Prediction

  • Predicting house prices
  • Estimating insurance premiums
  • Forecasting product costs

Forecasting and Trends

  • Predicting future sales revenue
  • Estimating energy consumption
  • Forecasting website traffic

Measurements and Quantities

  • Predicting delivery time
  • Estimating fuel efficiency
  • Calculating demand levels

All of these scenarios involve predicting a numeric value, making them regression problems.

Regression vs Other Machine Learning Techniques

Understanding the difference between regression and other ML techniques is critical for AI-900.

TechniqueOutput TypeExample
RegressionNumeric valuePredicting house price
ClassificationCategory or labelApproving or denying a loan
ClusteringGroup assignmentSegmenting customers
Anomaly detectionUnusual behaviorDetecting fraud

Exam tip:
“Yes/No”, “True/False”, or named labels → Classification
A number or measurement → Regression

Example Exam Scenarios

Scenario 1

A company wants to predict the monthly electricity usage of buildings based on historical data.

  • Output: Electricity usage (kWh)
  • ML Technique: Regression

Scenario 2

A real estate company wants to estimate the selling price of homes based on size, location, and age.

  • Output: Price
  • ML Technique: Regression

Scenario 3

A logistics company wants to estimate delivery time for packages.

  • Output: Time (hours or days)
  • ML Technique: Regression

Azure Context for AI-900

On the AI-900 exam, regression scenarios are often framed using Azure Machine Learning concepts:

  • Training models using historical datasets
  • Predicting numeric outcomes
  • Evaluating prediction accuracy

You are not expected to:

  • Write code
  • Choose algorithms
  • Tune hyperparameters

Focus on recognition, not implementation.

Common Exam Traps and Misconceptions

  • ❌ Predicting categories like high / medium / low → Classification
  • ❌ Grouping similar items without labels → Clustering
  • ❌ Detecting rare events → Anomaly detection
  • ✅ Predicting a number → Regression

Key Takeaways for the Exam

  • Regression predicts numeric values
  • It is a supervised learning technique
  • Look for words like predict, estimate, forecast
  • Outputs are continuous values, not categories
  • Regression is commonly used for prices, quantities, and time

Go to the Practice Exam Questions for this topic.

Go to the AI-900 Exam Prep Hub main page.

Leave a comment