2 ms·

Clustering is a poorly defined task. What constitutes a cluster, separation by euclidean distance? That's not scale invariant. The representation and clusterin

• binarymax 2y ago
  This is one of those interesting mathematical attempts to formalize a very human “I know it when I see it” quality. So maybe k-means is a poorly defined task if you don’t know k. Then you end up with the meta task of defining k, which has its problems, as seen in the paper. K-means alone, on some unknown data, without knowing the distance metric, and with no other heuristics or analysis, is going to go wrong. But if you are performing a dataset specific task, in a known vector space, with a good understanding of the outcome, then it’s really useful. The problem is probably that people learn that k-means is an unsupervised learning algorithm, and apply it incorrectly.
  • hoseja 2y ago
    Thing is, you only see it when appropriately zoomed in and therefore may not actually see it even though the algorithm should be able to find it.