Cluster Assigner is indifferent

Hey guys,

my Cluster Assigner seems to be indifferent. The situation:

I have training data (vectors) to perform a k-means. The clustering rules are provided to a cluster assigner. One vector of the test data seems to be exactly in the middle of some clusters. I get a "?" as clustering result. If I add 10^(-30) to all values of this vector, the clustering result is clear.

How can I prevent this from happening? Do I have to check each result for a "?" and assign it again if this happens?

Thank you very much for support!!

(if you need the data I will load them up)



Hi Jan,

see Thorsten answer in