There are a few more advanced arguments you can pr...
# datascience
t
There are a few more advanced arguments you can provide, like
k1
and
k2
. This throttles the "smoothing" so no item has a 0% chance to be labeled with a given category.