Majority vote in Ensemble Learning

This works better for the classification tasks. From all the prediction of the weak learners, we take the one with the majority votes and in case tie, randomly pick one.


Related Notes