Distance Functions Expansion

KNIME Analytics Platform
1

With the new Distance Functions process allowing this to be defined earlier with these nodes and subsequently applied to k-Medoids and Hierarchical Clustering, is there a possibility for this to be extended to other clustering techniques such as the following, assuming the algorithms for these are compatible that is;

- Fuzzy c-Means

- k-Means.

- SOTA Learner.

 

 Thanks

Simon.

2

Hi Simon, 

I am told these algorithms are not easily adaptable to work with distance matricies, but rumor has it that there is a DBSCAN implementation in the works which should be quite handy.

http://en.wikipedia.org/wiki/DBSCAN