Talk:Diffusion map

Tried to tidy this up as best I can, given my (very, very) limited understanding of the subject. Apologies if it no longer makes sense. DoctorKubla (talk) 22:01, 8 March 2012 (UTC)

Duplication

I did a bunch of cleanup edits on the article titled diffusion maps (plural), then I tried to change the title to diffusion map (singular) and was told that an article with that title already exists. It turned out that was (at least at a superficial inspection) nearly identical to the one with the plural title. Bizarre!! Why did people tolerate that situation? Anyway, I changed the title and then merged the two edit histories. Michael Hardy (talk) 01:42, 20 October 2012 (UTC)

What is this?

What is this article about? I can't understand it.

For example, what is a data set? Might it be a large ordered collection of points from some high dimensional Euclidean space?

What kind of manifold is the underlying manifold that the data has been sampled from -- is it meant to be a submanifold of some high dimensional Euclidean space?

Most mathematics articles on Wikipedia are written with sufficient technical precision and clarity that a mathematician can understand them -- in fact, the level is generally excellent. This one is not there yet. 84.227.254.183 (talk) 15:34, 9 September 2014 (UTC)

This page should be merged with Spectral clustering, they are the same technique just discussed differently. The other page frame the problem from the perspective of finding a cut in the graph which is an NP hard problem which is why people use the gaussian similarity and the Laplacian matrix instead. AngCampbell (talk) 14:48, 4 March 2021 (UTC)

Examples

The circular helix example should included and be fully worked out Izmirlig (talk) 12:08, 12 December 2024 (UTC)