Generalized K-Means Clustering • Hacker News Board

Re: Generalized K-Means Clustering

#11

Post by 3abiton » Mon, Oct 27, 2025, 1:39 AM UTC

dcl wrote: post #10

nurettin wrote: post #4

Earlier quoted context omitted.

> And so that raises the question of what “nearest” means To me, the definition of "nearest" is just a technicality. The real question is: what is K?

At a previous $dayjob at a very large financial institution, it's however many clusters are present in the strategy that was agreed to by the exec team and their highly paid consultants. You find that many clusters and shoehorn the consultant provided categories on to the k clusters you obtain.

To be fair finding K is highly domain dependent and I would argue should not be for the analyst (solely) to decide, but with a feedback from domain experts.

Re: Generalized K-Means Clustering

#12

Post by atiedebee » Mon, Oct 27, 2025, 3:56 PM UTC

apwheele wrote: post #7

Can folks comment on what applications they use k-means for? It was a basic technique I learned in school, but honestly I am not really familiar with a single use case that is very clearly motivated besides "pretty pictures". So I do a bit of work in geospatial analysis, and hotspots are better represented by DBSCAN (do not need to assign every point a cluster). I just do not even use clustering very often in gig (su…

It's used for vector quantization which can be used for color quantization

Re: Generalized K-Means Clustering

#13

Post by keeeba » Tue, Oct 28, 2025, 9:28 AM UTC

nurettin wrote: post #8

mentalgear wrote: post #5

Earlier quoted context omitted.

Have you tried HDBSCAN (DBSCAN variant) or Hierarchical Clustering (HAC) ?

Me? I probably tried every classification algorithm and their H variants. I still think "What is K?" is a profound question.

I agree it is a profound question. My thesis is fairly boring.

For any given clustering task of interest, there is no single value of K.

Clustering & unsupervised machine learning is as much about creating meaning and structure as it is about discovering or revealing it.

Take the case of biological taxonomy, what K will best segment the animal kingdom?

There is no true value of K. If your answer is for a child, maybe it’ 7 corresponding to what we’re taught in school - mammals, birds, reptiles, amphibians, fish, and invertebrates.

If your answer is for a zoologist, obviously this won’t do.

Every clustering task of interest is like this. And I say of interest because clustering things like digits in the classic MNIST dataset is better posed as a classification problem - the categories are defined analytically.