The operator [Stacking] and [Voting], according to operator description and sample process, sound no difference from each other. Anyone can tell me what exactly the difference of these two on usage?
Hi, in Stacking, one learner learns from the output of another. When you use voting, each learner is learned on the original data set and the class that is most often returned by the different models is used.
Actually I had also seen that the description of the Vote operator needed an adjustment, as currently the one from Stacking is used there, which may have created the confusion.
Hi, this is of course a problem. But since we will soon have a Wiki containing all operator descriptions, you can correct it yourself if you feel that there's an problem
Contributor II
Answers
in Stacking, one learner learns from the output of another. When you use voting, each learner is learned on the original data set and the class that is most often returned by the different models is used.
Greetings,
Sebastian
Actually I had also seen that the description of the Vote operator needed an adjustment, as currently the one from Stacking is used there, which may have created the confusion.
Regards,
Dan
this is of course a problem. But since we will soon have a Wiki containing all operator descriptions, you can correct it yourself if you feel that there's an problem
Greetings,
Sebastian