Microsoft Azure AI Fundamentals (AI-900) Practice Exam

Which scenario is most likely NOT a use case for clustering?

1. Identifying customer segments based on purchasing behavior

2. Predicting future stock prices

3. Organizing similar news articles into categories

4. Grouping similar items for recommendations

The correct answer is: Predicting future stock prices

Clustering is a type of unsupervised machine learning technique primarily used for grouping similar items based on their characteristics without prior knowledge of the labels. In the context of the options provided: Identifying customer segments based on purchasing behavior involves analyzing and categorizing customers with similar buying habits, which is a classic use case for clustering. Organizing similar news articles into categories also fits well within clustering as it requires grouping articles that share similar topics or themes. Grouping similar items for recommendations is directly aligned with clustering as it seeks to identify groups of items that appeal to similar user preferences, enhancing personalized recommendations. Predicting future stock prices, however, is more aligned with regression analysis or time series forecasting, where historical data is used to make predictions about future values. This scenario does not involve clustering, as it is not about grouping data points based on similarity but rather making numerical predictions based on existing trends and patterns. Therefore, it is the scenario that most likely does not represent a use case for clustering.