How to select data for random forest classification?

Question

Algorithms

Hello!
In general, I need to study classification random forest.
It seems that the method is partially clear. And what about the data. They must be initially divided into classes, or are they divided into classes on their own? Then what is the algorithm? And what is the optimal number of classes?

Answer 1

Are you sure you studied theory?
Classification tasks (everything, regardless of the method) - they take as input a MARKED data set, on which a classifier is built, and in your case - a tree or forest. Then the resulting classifier is used to assign newly incoming data to one of the classes.
"Independent" (in your interpretation) division into classes is another task - clustering.