To create a hierarchy of clusters from a dataset

To perform dimensionality reduction

To create a single partition of the dataset

To identify the principal components