As I know the experimenter allows to compare the performance of two or
more classifiers based on statistical tests .
The question is how we evaluate the performance of same classifier on two
data sets in Weka based on statistical tests as in the experimenter . For
example , evaluation whether there is a significant different in the
performance before feature selection and after feature selection.
View this message in context: http://weka.8497.n7.nabble.com/Statistical-tests-to-evaluate-a-classifier-t…
Sent from the WEKA mailing list archive at Nabble.com.