Courseiva

AI-900 Practice Question: Describe fundamental principles of machine learning on Azure

A data scientist wants to train a model that predicts whether a customer will respond to a marketing offer (yes or no). The dataset includes features such as age, income, past purchase history, and the labeled outcome (responded or not responded) for previous customers. Which type of machine learning is this?

⚠ Common exam trap

The trap here is that candidates might confuse supervised learning with unsupervised learning, thinking that because the dataset has many features (age, income, etc.) it must be unsupervised clustering, but the presence of labeled outcomes clearly indicates supervised classification.

Answer choices

Why each option matters

Answer the question above first, then reveal the full breakdown to understand why each option is right or wrong.

Correct answer & explanation

Supervised learning

This is supervised learning because the dataset includes labeled outcomes (responded or not responded) for previous customers, which the model uses to learn a mapping from input features (age, income, past purchase history) to the correct output. The goal is to predict a categorical label (yes/no), making it a classification task within supervised learning.

Answer analysis

Option-by-option breakdown

For each option: why learners choose it and why it is or isn't the right answer here.

  • Supervised learning

    Why this is correct

    This is a supervised learning task because the training dataset consists of historical customer records where each instance has both predictor features (e.g., usage, tenure, demographics) and a known ground-truth label indicating the outcome. The model optimizes a loss function by comparing its predictions against these true labels, and because the label is a discrete category (churn vs. no churn), it is specifically classification rather than regression.

  • Unsupervised learning

    Why it's wrong here

    Unsupervised learning is not appropriate here because it assumes the input data has no target labels and the goal is to infer hidden structure, such as customer segments via clustering (e.g., k-means) or low-dimensional representations via PCA. In this scenario, every prior customer already has an observed churn outcome, so the task is to generalize from labeled examples rather than discover patterns absent a target variable.

  • Reinforcement learning

    Why it's wrong here

    Reinforcement learning does not fit this problem because it requires an agent interacting with an environment over time, taking actions and receiving rewards or penalties that guide policy learning. The prediction task here involves a fixed, static dataset of customer attributes and outcomes, with no sequential decision-making, no environment transitions, and no reward signal to maximize, so RL's trial-and-error framework is irrelevant.

  • Semi-supervised learning

    Why it's wrong here

    Semi-supervised learning is incorrect because it is designed for situations where only a small fraction of the available data carries labels and the rest is unlabeled, often used when labeling is expensive. The problem statement explicitly states that all previous customers have labeled outcomes, so there is no set of unlabeled instances to exploit; this removes the defining condition of the semi-supervised paradigm.

About these practice questions

This AI-900 question is part of Courseiva's 985-question bank — original exam-style content with full explanations and wrong-answer analysis, never real exam questions or exam dumps. Learn why practice questions differ from exam dumps →

How Courseiva writes practice questions · Editorial policy

JA

Written by Johnson Ajibi, MSc IT Security

Senior Network & Security Engineer · founder of Courseiva

This AI-900 practice question is part of Courseiva's free Microsoft certification practice question bank. Courseiva provides original exam-style practice questions with explanations, topic-based practice, mock exams, readiness tracking, and study analytics to help learners prepare for the AI-900 exam.