Unsupervised Learning: Algorithms and Business Use Cases

Understanding customers is important for businesses. Usually different groups of users can be split across some criteria or into different groups. The criteria or grouping can be as simple as age and gender or as complex as personality type and purchase history.

Clustering algorithms are able to go through the data and find these natural clusters if they exist; they can segregate groups with similar traits and assign them into clusters. For example, clustering might create one cluster of retired people who own a pet and another of millennials who play guitar. We can typically define the number of clusters we want; this allows us to adjust the granularity of these groups.