AWS Certified AI Practitioner (AIF-C01) — Question 345

A company wants to build an ML model to detect abnormal patterns in sensor data. The company does not have labeled data for training.

Which ML method will meet these requirements?

Answer options

• A. Linear regression
• B. Classification
• C. Decision tree
• D. Autoencoders

Correct answer: D

Explanation

Autoencoders are an unsupervised learning technique well-suited for anomaly detection because they can learn to reconstruct normal data patterns and flag instances with high reconstruction errors as anomalies. In contrast, linear regression, classification, and decision trees are supervised learning methods that require labeled target variables for training, making them unsuitable for unlabeled sensor data.