cluatering of large network

Legacy User

Hello -
    Anyone used RapidMiner for clustering of large network (100K nodes)? Looking for advices of what to watch out for a newbie.

Thx

Legacy User

Hi,

I have just worked on 120000 data points today (5 dimensions, K-Means clustering with K = 30) and it took about half an hour.

However, I am not sure if this also holds for version 4.2 since I use the CVS developer version of RapidMiner. And the guys from Rapid-I seem to have drastically optimized the clustering schemes with respect to runtime.

Best regards,
Peter