With unsupervised clustering, we aim to determine "natural" or "data-driven" groups in the data without using apriori knowledge about labels or categories. The challenge of using different unsupervised clustering methods is that it will result in different partitioning of the samples and thus different groupings since each method implicitly impose a structure on the data. Thus the question arises; What is a "good" clustering? Figure 2A depicts a bunch of samples in a 2-dimensional space. Intuitively we may describe it as a group of samples (aka the images) that are cluttered together. I would state that there are two clusters without using any label information.
Dec-11-2021, 02:55:18 GMT