A4.3.4 Unsupervised Learning: Clustering
Every technique so far has been supervised: the data came with labels, and the model learned to predict a known answer. Clustering is different. It works on data that has no labels at all, and its task is to discover the natural groups hiding in that data on its own. This page describes what clustering does, walks through k-means as the standard example of how a clustering algorithm groups data by feature similarity,...