A team wants to classify support tickets as positive/negative/neutral using a dataset of historical tickets WITH labeled outcomes. Which ML approach?
A.Supervised learning — classification on labeled data
Labels + categories = supervised classification. Correct.
B.Unsupervised learning — clustering unlabeled data
No labels are used in unsupervised — you HAVE labels here.
C.Reinforcement learning
RL is for sequential decision problems — not text labeling.
D.Self-supervised pre-training
Pre-training builds base models — not the direct classification answer.
Why this answer
Labeled historical data + a target category = supervised classification, the canonical AIF exam mapping. Unsupervised finds structure WITHOUT labels; reinforcement learns by rewards; self-supervised builds representations for pre-training.
Correct answer: A · Last reviewed 2026-11-05
Official AWS references
Practice this interactively
Answer without seeing the solution first in the AI/ML Foundations practice session.
Practice AI/ML Foundations