One of the most common ways to apply unsupervised learning to a dataset is clustering, specifically centroid-based clustering.