CNB is good for multi-class classification in which a single label is applied from several options. This is different from multi-label classification in which several labels are applied.
On Fri, Mar 1, 2013 at 1:29 AM, Sreejith S <[email protected]> wrote: > Hi All, > > While searching the mahout mailing list i found a comment that > "Complementary Naive Bayes" is perfect for multi-label classification. > > Can any one give some pointers how to achieve this ? Its better to get > some hints on how to go further. > > > Thank you, > > -- > > > *Sreejith.S* >
