I am currently working on a multi-class classification problem using BDTG (Boosted Decision Tree Gradient). I would want to add the probability of a classifier per event. For example, I would want my classifier to say: “There is a 80% probability that this event belongs to class B”. How do I do that?
You would determine the classifier performance on a given sample. That tells you its efficiency and mis-classification rate for events similar to that sample.