I would like to use the most intuitive method like minimizing the within-cluster distance and maximizing the distance between neighbouring clusters, but not sure does this method have a name or something? Is there any related paper regarding it? The only method I'm using now is Silhouette score. Many thanks.

  • 1
    $\begingroup$ It's not clear if your question is about clustering evaluation or clustering methods? For proper evaluation one needs annotated data and/or a task-specific method to measure the quality of the clusters. $\endgroup$ – Erwan Jul 31 '19 at 23:57
  • $\begingroup$ What is wrong with Silhouette? It is too intuitive and too little mathematically based in my opinion, I would have defined it differently... $\endgroup$ – Has QUIT--Anony-Mousse Aug 1 '19 at 5:27
  • $\begingroup$ @Anony-Mousse Hi thanks for the reply, nothing's wrong with it, just want to know if there is any other evaluating method. In terms of Silhouette, I followed this tutorial:scikit-learn.org/stable/auto_examples/cluster/…, but I'm a bit confused about the concept, is it only consider the distance between one point and neighbouring clusters, so the score is the higher the better What about the within cluster distance which we want it the smaller the betterʔ $\endgroup$ – Cecilia Aug 1 '19 at 9:34
  • $\begingroup$ Silhouette uses both distances and scores high if they differ, i.e. within << to next other cluster. Make sure you understand the a and b terms. $\endgroup$ – Has QUIT--Anony-Mousse Aug 1 '19 at 23:34
  • $\begingroup$ Thanks @Anony-Mousse $\endgroup$ – Cecilia Aug 2 '19 at 7:35

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.