While reading about ensembles I came across something which I didn’t quite understand.
What does the last bullet point mean by fitting a model on the probabilities??
Don’t the models calculate the probabiities??How can a model be trained on probabilities??
Can someone please clear these doubts??