I explain how we can determine an upper bound to the error of a strong classifier using AdaBoost.
No download links available.