Semisupervised Learning

cpc2

Hi,
I would like to use the NaiveBayes Op. to learn classes from a labeled dataset and then extend it to a unlabeled dataset using the EMClustering Op.. Is there any way how i can do this in a single process or do I have to split this into two processes ? ( And if this can be done in one process how ? )
Thanks in advance, Birger

fischer

Hi,

to be honest, I have no idea what you are trying to achieve. In general, anything that can be done in two processes in RM can also be done in one. Just copy your input using an IOMultiplier, consume everything you don't need using an IOConsumer at the end of the first process and then execute the second process. You may want to group subprocesses into OperatorChains for clarity.

If you need particular help concerning your process, you have to be more precise. What do you mean by extend? What do you want to extend? The model? The example set? Maybe you want to apply your NaiveBayes model to unlabelled data using a ModelApplier? Please provide more information.

Best,
Simon