-2

I have bunch of data.I am clustering this data with hierarchical clustering algorithm. How can i find the probability that an instance belongs to a cluster. When I found probability, I will make some calculations.Then, I will add this instance my train set according to result.

Has QUIT--Anony-Mousse
  • 76,138
  • 12
  • 138
  • 194
Var Yok
  • 1
  • 3

2 Answers2

0

Hierarchical clustering does not compute a probability. It is not a probabilistic model - it does not provide probabilities.

So you will have to come up with your own modeling approach, and I don't think it will be easy to retrofit a probability model onto this method.

Has QUIT--Anony-Mousse
  • 76,138
  • 12
  • 138
  • 194
0

If you want to add a new column with the cluster assignments to your data set, click "Filter" / "Unsupervised" / "attribute" / "AddCluster" (Then parameterize the clustering algorithm) / "Apply Filter". This will add the Cluster ID, that Weka thinks is appropriate for this parameterization, as a new column in your dataset.

But this won't give you any "probabilities of cluster assignments" because that's -as user Anony-Mousse remarked- not how hierarchical clustering works.

knb
  • 9,138
  • 4
  • 58
  • 85