I am having some difficulty understanding Adaboost.

How should the 1st threshold/classifier/weak learner be chosen?

It seems that there are two conditions which must be satisfied

  1. Choose the classifier with the lowest error
  2. $e(t)<0.5$ otherwise stop;

But if condition 1 is satisfied, doesn't it imply that condition 2 will also be satisfied automatically?


1 Answer 1


Yes, this is simply to ensure that all of the weak learners in the final solution classify the training data better than chance (which is essential for the theoretical properties of AdaBoost to hold).


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.