K-means and “Use static random seed“

KNIME Analytics Platform
1

Hi, just a quick question regarding the k-means node in KNIME. What exactly does it mean to enable “Use static random seed” in the configuration of the node? For example, what does it mean if I write 5 … 100 next to the initialization?

I’m trying to figure out whether I should have run my k-means node several times with different starting centroids to find the most optimal distortion - and thus get clusters that, in my case, best represent different user groups on a website.

Thank you in advance!

2

Hi @Ida0347.

Because the initial placement of centroids is random, different runs can produce different results. Setting a static random seed ensures that the random number generator produces the same sequence of initial centroids each time, making the results reproducible.

Br

3 Likes