2
$\begingroup$

I would like to have some suggestions on possible avenues that would make sense in the following context.

  • 3 Optimal clusters have been identified in a 5000 list of customers using Kmeans
  • Data model has 30 features and a PCA was performed prior to Kmeans.

I would like to further breakdown each of these 3 clusters into smaller tiers for each cluster. These tiers would server in ranking each customer within his cluster.

For example: Cluster 1, 2, 3 could all have 4 or 5 tiers.

What interesting option could I explore to create the tiers within the 3 clusters (e.g. RFM quantile separation, classification algorithms)?

$\endgroup$
3
  • $\begingroup$ why not kmeans again? $\endgroup$
    – Nikos M.
    Nov 9, 2021 at 17:53
  • $\begingroup$ maybe hierarchical clustering can cluster and subcluster automatically $\endgroup$
    – Nikos M.
    Nov 9, 2021 at 17:53
  • 1
    $\begingroup$ @NikosM. i've tried kmeans but it isnt performing in a desirable way. hierarchical could be interesting $\endgroup$ Nov 9, 2021 at 22:14

1 Answer 1

0
$\begingroup$

One option is hierarchical clustering which builds a hierarchy of clusters.

$\endgroup$

Your Answer

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

Not the answer you're looking for? Browse other questions tagged or ask your own question.