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

Answer analysis

Option-by-option breakdown

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

• ✗

  Classification

  Why it's wrong here

  Classification is a supervised learning task that predicts a discrete class label from a fixed set of categories, such as 'spam' or 'not spam'. It is inappropriate here because the desired output is a continuous numeric count (the number of bike rentals), not a category. Even a multi-class classifier would fail to represent all possible rental counts and would discard the natural ordering and magnitude of the numeric target.

• ✓

  Regression

  Why this is correct

  Regression is the correct supervised learning technique because the target variable is a continuous numeric value—the number of bike rentals. A regression model learns a mapping from input features to a real-valued output, allowing it to predict rental counts that were not present in the training data. Common regression algorithms include linear regression, decision trees, random forests, and neural networks with a linear output unit, all of which can output a continuous number.

• ✗

  Clustering

  Why it's wrong here

  Clustering is an unsupervised learning method that groups similar data points into clusters based on feature similarity, with no labeled target variable. Algorithms such as k-means or DBSCAN assign each instance to a group, but they do not produce a numeric prediction for a specific rental count. Because this problem has historical labels (actual rental numbers) to learn from, clustering cannot be used to make supervised continuous predictions.

• ✗

  Anomaly Detection

  Why it's wrong here

  Anomaly detection is designed to identify rare, unusual, or outlier data points, such as fraudulent transactions or equipment failures, rather than to predict a typical continuous value. It is often implemented using one-class classification or novelty detection methods, which learn a boundary around normal behavior. In this scenario, anomaly detection might flag unusually high or low rental days, but it would not output the expected number of rentals for a given set of conditions, so it is the wrong task.

Go deeper

Related to this question