Let's say I have a few customers buying product A. How can I find other customers that have the same characteristics, or something close to, these A buyers?

I was thinking about using clustering methods, but I don't know if it's possible to "lock" some characteristic (for instance, putting all customers that have the locked characteristic in the same cluster).

To sum up, the question would be: is there any way of finding similar customers to a pre-selected group?

  • $\begingroup$ Couldn't you just run kmodes? $\endgroup$
    – dekio
    Apr 11, 2023 at 15:59

1 Answer 1


One way is nearest neighobor search. "Nearest" has to be defined, typically through a distance metric. Distance metrics depend on how the data is measured. If the data is measured continuously then Euclidean distance can be used. Often customer data is categorical. For categorical data, Jaccard distance can be used. There are many other distance metrics.


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

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