"SVM Weighting"

bx01z

Alright,

I able to process text so that it is classified via SVM; however, I receive a message in the model that states that classification could only be done on two classifiers and I need to apply SVMWeighting for additional classifiers.

What operator is this referring to?
Do I add this before or after the model creation?

Thanks,
Bob

land

Hi Bob,
you can ignore this message unless you want to get a weighting of the attributes used for classification. These weights give you an impression how important attributes (that probably represent words in your case) are for the outcome of the classification.
But the SVM learner isn't able to perform a classification with more than two classes and calculate these weights at the same time. If you need the weights in such a multiclass problem, you should refer to the weight by SVM operator.

Greetings,
  Sebastian