There is a pretty easy command line interface for doing this. Instantiating a Naive Bayes classifier from code is kind of complicated, but can be doped out by reading and hacking the code.
The SGD classifiers are designed to be easier to integrate on an API basis and Robin has been working on pulling something together for the Naive Bayes code. On Mon, Nov 22, 2010 at 7:13 PM, ivek gimmick <[email protected]> wrote: > I was able to test this example. But how do I use the actual classifier? > Once I train the data and have the model, I want to use the model to > categorize new set of data which is not classified. > > Is there any straight-forward way to do this with Mahout or should I be > tweaking the code? >
