Hi everybody,
I'm new in this forum and maybe the topic has already been discussed.
Anyway, I'd like to know how WEKA AddClassification filter works: i.e.: in order to predict whether a case is, say, of class A or B, which "test option" is used? is the training set itself used? does it perform a cross validation test?
If a tree model is used, is it possibile to display it somewhere?
Is there a way of getting accuracy rate, confusion matrix etc... automatically or should they be computed manually by comparing true class values and predicted ones?