"multiple learners in AdaBoost"

Elisa0815

I have a question about the AdaBoost operator.

As I understood AdaBoost in several literature, AdaBoost uses, for every iteration, another simple base classifier for classification. But in RapidMiner Studio, only one learner can be chosen for AdaBoost.

Did I may understand AdaBoost wrong? Why can only one learner be chosen? That doesn't fit to my understanding of AdaBoost and confuses me. ???

Would be great, if anyone could may help.

Greets
Elisa

MartinLiebig

Well, it is another instance of the same base learner.

So you always train e.g. a Decision tree but on differently weighted data. So the result are different decision trees.

~Martin