All of the examples that I've found for training NB classifiers seem
to have textual data as input.  Is there a way to build a classifier
with more general attributes?

I found this jira ticket
(https://issues.apache.org/jira/browse/MAHOUT-286), but it's been
closed:duplicate under
https://issues.apache.org/jira/browse/MAHOUT-155, which doesn't seem
to address the underlying question.

I know that I can do this with weka, but not at scale -- is mahout
only able to build textual classifiers?

Thanks,
Eric

Reply via email to