Considering there are multiple attributes.
Is there a way of saying this cluster is related to another in terms of so and so ?
How to determine confidence interval or with what probability that this member lies in a cluster?
Hi Sai, are the clusters already defined? If so you could compute the cross-cluster entropy or variation of information (link bellow). If you want an easy means of calculating cluster membership probabilities during clustering you could use fuzzy k-means.
http://link.springer.com/chapter/10.1007%2F978-3-540-45167-9_14
You could also use the Davis Bouldin Index (DBI Index), to measure the clustering performance. Some packages come with the DBI Index functions, I think MATLAB should have the function, and packages like Rapidminer have the DBI Index function inbuilt. http://en.wikipedia.org/wiki/Davies%E2%80%93Bouldin_index
Dear Sai,
Maybe Silhouette plots can help you:
Rousseeuw, Peter J. "Silhouettes: a graphical aid to the interpretation and validation of cluster analysis." Journal of computational and applied mathematics 20 (1987): 53-65.
Kind regards,
Olivier PARISOT
- Badges
- Science topic
- Similar topics
- Mathematics
- Statistics
- Probability