-1
$\begingroup$

Have say 1000 items and their pairwise correlations with each other, so 499,000 correlations. Want to use this correlation data to categorise the 1000 items into a limited number of categories that is set up front so say into 10 categories. Categories should work so that items in same category have relatively high correlation. What is a good way to approach/solve this problem?

$\endgroup$
1
$\begingroup$

There are many clustering algorithms that can be used with correlation distance.

In particular, try hierarchical clustering, DBSCAN, and OPTICS.

$\endgroup$

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.