Partitioning the dataset by locating observations based on centroids (center points of clusters) is called what?

Partitioning the dataset by locating observations based on centroids (center points of clusters) is called what?

• A. Data Clustering
• B. Data Transformation
• C. Principal Components Analysis
• D. Partitional Clustering

Answer: (D)

This is about forming groups by using cluster centers to assign observations. When you partition data by identifying center points — centroids — and assigning each observation to the nearest centroid, you’re applying partitional clustering. This approach is non-hierarchical and creates disjoint clusters, typically refined by re-computing centroids after each assignment (as in k-means). Data clustering is the broader idea of grouping similar data, but the term here points to the specific method that partitions around centroids. Data transformation and principal components analysis are different: they change the data representation or reduce dimensionality rather than partition it into clusters.

This is about forming groups by using cluster centers to assign observations. When you partition data by identifying center points — centroids — and assigning each observation to the nearest centroid, you’re applying partitional clustering. This approach is non-hierarchical and creates disjoint clusters, typically refined by re-computing centroids after each assignment (as in k-means).

Data clustering is the broader idea of grouping similar data, but the term here points to the specific method that partitions around centroids. Data transformation and principal components analysis are different: they change the data representation or reduce dimensionality rather than partition it into clusters.