True or false: The k-Means clustering is a non-predictive modeling technique.

• A. True
• B. False
• C. Not determinable
• D. Both true and false

Answer: (A)

k-Means is a clustering method, which is unsupervised learning. In predictive modeling, you aim to predict a target variable from labeled examples, but k-Means learns from data without any outcome to predict. It creates groups (clusters) by assigning each observation to the nearest centroid and updating those centroids to minimize within-cluster variance, providing a descriptive partitioning of the data rather than a function to predict a response. Even though you can place new data into an existing clustering by assigning it to the nearest centroid, that use is still applying a clustering result, not building a predictive model for a target variable. Therefore, it is considered non-predictive.