An implementation of adaboost.
Adaboost is one of the emsenble learning algorithms. The weak learner is a decision stump and the dataset used is "digits" imported from scikit-learn. In order to assess its performance, the error rate of the weak learner and the strong learner (that you constructed using adaboost) on each learning step can be calculated.