Let's say we take a quadtree of 50 dimension and apply dimension reduction(Assume the dimension reduction works well). Why does the dimension reduction not influence the clustering with DBSCAN(Mintpts:4 Ep:0.2), so the point in the clusters before reduction are likely the same after the reduction.

  • 1
    $\begingroup$ I suspect we're missing some context. Is this a question you saw somewhere? If so, can you tell us where you saw it? Why do you think the dimension reduction doesn't influence clustering with DBSCAN? What does the quadtree have to do with anything? What are your thoughts? What's the motivation for this question and why would one care? $\endgroup$
    – D.W.
    Jul 2 '21 at 20:02

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Browse other questions tagged or ask your own question.