Slicing the Rainbow: Clustering, k-means, DBSCAN, Mixtures and Segmentation Pitfalls

Deep End Era. Engineering with Sankalp · Oct 28, 2026 · 51m

Noor's seven customer segments became persona posters, until Rafael reran her code with a different random start and got seven different ones. Part five of The Deep End's Machine Learning series is about clustering, the problem with no answer key. We meet Weldon and Pearson's Naples crabs, Stuart Lloyd's quantiser that waited twenty-five years for publication, and three gyms that move every night until nobody switches. We read a family tree of languages as a dendrogram, find crowds and loners at a festival with DBSCAN, give customers soft memberships with Gaussian mixtures, and find…