Practice Exam Questions

---

Question 1

A bank wants to determine whether a credit card transaction is fraudulent.

Which machine learning technique should be used?

A. Regression
B. Classification
C. Clustering
D. Anomaly detection

Correct Answer: B

Explanation:
The output is Fraud / Not Fraud, which is a category. Predicting categories is a classification task.

---

Question 2

An organization wants to predict whether a customer will renew their subscription.

Which type of machine learning problem is this?

A. Regression
B. Classification
C. Clustering
D. Recommendation

Correct Answer: B

Explanation:
The outcome is Yes / No, which makes this a binary classification scenario.

---

Question 3

Which of the following scenarios is best suited for classification?

A. Predicting the price of a product
B. Grouping customers based on behavior
C. Determining if an email is spam
D. Estimating delivery time

Correct Answer: C

Explanation:
Spam detection involves assigning emails to Spam or Not Spam categories, which is classification.

---

Question 4

An AI system categorizes customer support tickets into predefined issue types.

What type of machine learning technique is being used?

A. Regression
B. Classification
C. Clustering
D. Time-series forecasting

Correct Answer: B

Explanation:
The system assigns each ticket to a known category, which is classification.

---

Question 5

Which output value most clearly indicates a classification scenario?

A. 128.5
B. 4.2 hours
C. High risk
D. 99.7

Correct Answer: C

Explanation:
High risk is a label, not a numeric value, indicating classification.

---

Question 6

A model predicts whether a customer will default on a loan.

Which machine learning approach is most appropriate?

A. Regression
B. Classification
C. Clustering
D. Anomaly detection

Correct Answer: B

Explanation:
Default / Not Default is a binary label, making this a classification problem.

---

Question 7

Which scenario represents multi-class classification?

A. Predicting house prices
B. Detecting unusual network traffic
C. Assigning images to animal types
D. Grouping products by sales patterns

Correct Answer: C

Explanation:
Assigning images to multiple animal types (cat, dog, bird) is multi-class classification.

---

Question 8

A healthcare system predicts whether a patient is at low, medium, or high risk.

Which type of machine learning is being used?

A. Regression
B. Classification
C. Clustering
D. Forecasting

Correct Answer: B

Explanation:
Low / Medium / High are categories, not numeric values, so this is classification.

---

Question 9

Which statement best describes classification models?

A. They predict continuous numeric values
B. They group unlabeled data
C. They assign inputs to predefined categories
D. They detect rare anomalies

Correct Answer: C

Explanation:
Classification models assign data points to predefined labels or categories.

---

Question 10

On the AI-900 exam, which keyword most strongly indicates a classification scenario?

A. Forecast
B. Estimate
C. Categorize
D. Measure

Correct Answer: C

Explanation:
Categorize indicates assigning labels, which is classification.

---

Exam-Day Tip

For machine learning related questions, if the question describes …

• Yes / No decisions
• Named labels
• Risk levels or categories

… the correct answer is likely related to Classification.

---

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