I'm reading about the bagging systems. After creating M data set using Bootsrapping we should train base models on these M data sets. How about base models? In Weka when I select bagging classifier I can select just one base learner. Does that mean in Bagging we should use just one base model??
Thanks in advanced