In Adaboost, how is it determined what is correct and what not? In the following example from StatQuest (in youtube), what correct is and what incorrect makes sense in real life. But what if we have a problem where we doesn't know that? What if:
- under 'weight>176'->'true' we change it to 'no heart disease'? We will have correct=0 and incorrect=3
- under 'weight>176'->'false' we change it to 'yes heart disease'? We will have correct=1 and incorrect=4
