Comment by xnickb

So you are saying we need to conduct trials on a representative set of bayesnets of say.. 200k and then it'll be conclusive?