On Jan 4, 2012, at 7:59 PM, Lance Norskog wrote:
> 
> Is it worthwhile to clamp the training data so that there are similar
> numbers of documents for each label? Or does Naive Bayes work well
> with a bell curve?

It's possible to do this already with the --maxItemsPerLabel.  I think I use 
that somewhere in there for NB setup.

Reply via email to