Home > Datasets, experiments, what i've been doing > Updated Hierarchy Stats

Updated Hierarchy Stats

February 16th, 2012 Leave a comment Go to comments

It turns out that the two clumps for Studivz-three-B related to K=3 and K=4. So we ran more runs on Enron and Mit to see what happens, the following are only for K=3.

In the case of the MIT datasets, I included the 4 thresholds we have been using (20th,80th,mean and median thresholds)  and a range of 20 values from the lowest edge weight to the highest. I will also run this on the Enron Subsets.

 

MIT-OCT

MIT-NOV

MIT-ALL

Enron (all)

  1. No comments yet.
  1. No trackbacks yet.