1
$\begingroup$

In Weka, there is a clustering algorithm with the name as Make Density Based Clusterer. When going through its properties, it takes a clusterer as base clusterer(I took it as K-means with k=3). It initially performs k-means and creates three clusters. I see prior probabilities for each cluster and attribute-wise normal distribution means and standard deviation in the result buffer.

What happens after k-means clusters are calculated? What role mean, standard deviation and prior probabilities play here? Why is it called density based?

$\endgroup$
1
$\begingroup$

Based on this paper, MakeDensityBasedClusterer is a metaclusterer that wraps a clustering algorithm to make it return a probability distribution and density. To each cluster and attribute, it fits a discrete distribution or a symmetric normal distribution (whose minimum standard deviation is a parameter).

$\endgroup$
1
  • $\begingroup$ How does this result in re-distribution of points across the clusters obtained from wrapped clustering algorithm(int this case K-means) ? $\endgroup$ Mar 18 '19 at 6:07

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.