Can a Bayesian filter be used to create multiple outputs

I see that Bayesian filters are use well for binary choices - (spam:not spam, male:female etc). Is there any way for it to categorize multiple values (eg php+javascript, house+yard). I've seen Naive bayesian classifier - multiple decisions but I want to know if multiple outputs are possible.

If not, what are other suggested approaches for categorization (with or withou开发者_如何转开发t learning). Especially for php.

As the accepted answer of the question you linked to says: "It's definitely possible to have more than two classes.". In practice, one approach is to train multiple classifiers in parallel, e.g. one classifier for php vs. not php and another classifier for javascript vs. not javascript.

Other widely used multivariate classification methods include

• artificial neural networks (also called multilayer perceptrons)
• (boosted) decision trees
• support vector machines

If you have a more detailed/follow up question on this, post it on http://stats.stackexchange.com .

I'm not sure what libraries for such a task are available for php but Swig is a tool to make libraries written in C/C++ usable from php.