Unsupervised machine learning deals with unlabeled data. This type of learning can discover all kinds of unknown patterns in the data and can facilitate useful categorization. Consider a scenario where patients use an online web application to learn about a disease, learn about their symptoms, and manage their illness. Such web applications that provide psychoeducation about certain diseases are referred to as Internet-Delivered Treatments (IDT). Imagine several thousand patients accessing the website at different timestamps of the day, learning about their illness, and all their activities are being logged into our database. When we analyze these log files and plot them using a scatter plot, we find a large group of patients who are accessing the website in the afternoon and a large chunk accessing the website in the evening. Some other patients also follow random login patterns. This scenario illustrates two distinct clusters of patients: one active in the afternoon and one active in the evening. This typical scenario is an example of a clustering task. 

There are several types of unsupervised learning algorithms that we can use. However, two major unsupervised learning tasks are clustering and dimensionality reductions. In the next section, we will discuss more regarding the different applications of unsupervised learning algorithms.